def test_standardizer_different_configs(self):
"""Testing standardization of smiles using threading"""
df_smiles = read_csv(curDir / "input/test_standardizer.csv")
outcols = ["canonical_smiles", "success", "error_message"]
out_types = ["object", "bool", "object"]
## Load ref standardizer
st_ref = Standardizer(
max_num_atoms=self.config["standardization"]["max_num_atoms"],
max_num_tautomers=self.config["standardization"]
["max_num_tautomers"],
include_stereoinfo=self.config["standardization"]
["include_stereoinfo"],
verbosity=0,
)
dt_ref = DfTransformer(
st_ref,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=4,
verbosity=0,
)
response_ref = dt_ref.process_dataframe(df_smiles)[0]
config_2 = ConfigDict(config_path=Path(
os.path.join(curDir, "input/",
"example_parameters_2.json"))).get_parameters()
## load test standardizer
st_tmp = Standardizer(
max_num_atoms=config_2["standardization"]["max_num_atoms"],
max_num_tautomers=config_2["standardization"]["max_num_tautomers"],
include_stereoinfo=config_2["standardization"]
["include_stereoinfo"],
verbosity=0,
)
dt_tmp = DfTransformer(
st_tmp,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=2,
verbosity=0,
)
response_tmp = dt_tmp.process_dataframe(df_smiles)[0]
try:
assert_frame_equal(response_ref, response_tmp)
except AssertionError:
# frames are not equal
pass
else:
# frames are equal
raise AssertionError
def test_standardizer_parameter_atom_count(self):
"""Testing standardization with different number of max atom count"""
df_smiles = read_csv(curDir / "input/test_standardizer.csv")
outcols = ["canonical_smiles", "success", "error_message"]
out_types = ["object", "bool", "object"]
## Load ref standardizer
st_ref = Standardizer(
max_num_atoms=self.config["standardization"]["max_num_atoms"],
max_num_tautomers=self.config["standardization"]
["max_num_tautomers"],
include_stereoinfo=self.config["standardization"]
["include_stereoinfo"],
verbosity=0,
)
dt_ref = DfTransformer(
st_ref,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=4,
verbosity=0,
)
response_ref = dt_ref.process_dataframe(df_smiles)[0]
## load test standardizer
st_tmp = Standardizer(
max_num_atoms=5,
max_num_tautomers=self.config["standardization"]
["max_num_tautomers"],
include_stereoinfo=self.config["standardization"]
["include_stereoinfo"],
verbosity=0,
)
dt_tmp = DfTransformer(
st_tmp,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=2,
verbosity=0,
)
response_tmp = dt_tmp.process_dataframe(df_smiles)[0]
try:
assert_frame_equal(response_ref, response_tmp)
except AssertionError:
# frames are not equal
pass
else:
# frames are equal
raise AssertionError
def test_scramble_desc_multiple_key(self):
"""test if scrambling is depending on the input key"""
newKey = "melloddy_2"
tempFilePathFeat = curDir / "output/tmp/ecfp_feat_scrambled_new_key.csv"
df_smiles = read_csv(curDir / "input/chembl/chembl_23_example_T2.csv",
nrows=10)
dc = DescriptorCalculator.from_param_dict(
secret=newKey,
method_param_dict=self.config["fingerprint"],
verbosity=0)
outcols = ["fp_feat", "fp_val", "success", "error_message"]
out_types = ["object", "object", "bool", "object"]
dt = DfTransformer(
dc,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=1,
verbosity=0,
)
df_test = dt.process_dataframe(df_smiles)[0]
df_test.to_csv(tempFilePathFeat,
index=False) # write test fingperprints
result = filecmp.cmp(
"unit_test/output/test_calculate_desc_y2.csv",
tempFilePathFeat,
shallow=False,
)
self.assertEqual(result, False)
def test_calculate_desc_multiple(self):
tempFilePath = curDir / "output/tmp/ecfp_feat_multiple.csv"
df_smiles = read_csv(curDir / "input/chembl/chembl_23_example_T2.csv",
nrows=10)
dc = DescriptorCalculator.from_param_dict(
secret=self.keys["key"],
method_param_dict=self.config["fingerprint"],
verbosity=0,
)
outcols = ["fp_feat", "fp_val", "success", "error_message"]
out_types = ["object", "object", "bool", "object"]
dt = DfTransformer(
dc,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=1,
verbosity=0,
)
# df_ref = dt.process_dataframe(df_smiles)[0] #calculate reference fingperprints
# df_ref.to_csv("unit_test/output/test_calculate_desc_y2.csv", index=False) #write reference fingperprints
df_test = dt.process_dataframe(df_smiles)[0]
df_test.to_csv(tempFilePath, index=False) # write test fingperprints
result = filecmp.cmp("unit_test/output/test_calculate_desc_y2.csv",
tempFilePath,
shallow=False)
self.assertEqual(result, True)
def test_standardizer_multiprocessing(self):
"""Testing standardization of smiles using threading"""
df_smiles = read_csv(curDir / "input/chembl/chembl_23_example_T2.csv",
nrows=10)
st = Standardizer(
max_num_atoms=self.config["standardization"]["max_num_atoms"],
max_num_tautomers=self.config["standardization"]
["max_num_tautomers"],
include_stereoinfo=self.config["standardization"]
["include_stereoinfo"],
verbosity=0,
)
outcols = ["canonical_smiles", "success", "error_message"]
out_types = ["object", "bool", "object"]
dt_2 = DfTransformer(
st,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=2,
verbosity=0,
)
response2 = dt_2.process_dataframe(df_smiles)[0]
dt_4 = DfTransformer(
st,
input_columns={"smiles": "smiles"},
output_columns=outcols,
output_types=out_types,
success_column="success",
nproc=4,
verbosity=0,
)
response4 = dt_4.process_dataframe(df_smiles)[0]
assert_frame_equal(response2, response4)