def test_structure_to_composition(self):
coords = [[0, 0, 0], [0.75, 0.5, 0.75]]
lattice = Lattice([[3.8401979337, 0.00, 0.00],
[1.9200989668, 3.3257101909, 0.00],
[0.00, -2.2171384943, 3.1355090603]])
struct = Structure(lattice, ["Si"] * 2, coords)
df = DataFrame(data={'structure': [struct]})
stc = StructureToComposition()
df = stc.featurize_dataframe(df, 'structure')
self.assertEqual(df["composition"].tolist()[0], Composition("Si2"))
stc = StructureToComposition(reduce=True,
target_col_id='composition_red')
df = stc.featurize_dataframe(df, 'structure')
self.assertEqual(df["composition_red"].tolist()[0], Composition("Si"))
def _add_composition_from_structure(self, df, overwrite=True):
"""
Automatically deduce compositions from structures if:
1. structures are available
2. compositions are not available (unless overwrite)
3. composition features are actually desired. (deduced from whether
composition featurizers are present in self.featurizers).
Args:
df (pandas.DataFrame): May or may not contain composition column.
overwrite (bool): Whether to overwrite the composition column if it
already exists.
Returns:
df (pandas.DataFrame): Contains composition column if desired
"""
if (
self.structure_col in df.columns
and (self.composition_col not in df.columns or overwrite)
and self.composition_col in self.featurizers
):
if self.composition_col in df.columns:
logger.info(
self._log_prefix + "composition column already exists, "
"overwriting with composition from structure."
)
else:
logger.info(
self._log_prefix + "Adding compositions from structures."
)
df = self._tidy_column(df, self.structure_col)
# above tidy column will add oxidation states, these oxidation
# states will then be transferred to composition.
struct2comp = StructureToComposition(
reduce=True,
target_col_id=self.composition_col,
overwrite_data=overwrite,
)
struct2comp.set_n_jobs(self.n_jobs)
df = struct2comp.featurize_dataframe(df, self.structure_col)
return df