def test_StandardScaler_output(test_input):
scaler = Scaler()
scaler.execute(params=test_input)
if "Distance" in test_input["train_df"].keys():
assert round(test_input["train_df"]["Distance"][0], 5) == 1.08006
else:
assert round(test_input["train_df"]["Negatives"][0], 5) == 0.57771
def test_BinaryScaler_output(test_input):
scaler = Scaler()
scaler.execute(params=test_input)
assert (test_input["df"]["Negatives"].values.any() == 1
or test_input["df"]["Negatives"].values.any() == 0)
assert not (test_input["df"]["Negatives"].between(0, 1,
inclusive=False).any())
assert test_input["df"]["Negatives"][0] == 1
def test_MinMaxScaler_output(test_input):
scaler = Scaler()
scaler.execute(params=test_input)
if "Distance" in test_input["train_df"].keys():
assert test_input["train_df"]["Distance"].values.all() >= 0
assert test_input["train_df"]["Distance"].values.all() <= 1
else:
assert test_input["train_df"]["Negatives"].values.all() >= 0
assert test_input["train_df"]["Negatives"].values.all() <= 1
def test_BinaryScaler_output(test_input):
scaler = Scaler()
scaler.execute(params=test_input)
assert (test_input["train_df"]["Negatives"].values.any() == 1
or test_input["train_df"]["Negatives"].values.any() == 0)
assert not (test_input["train_df"]["Negatives"].between(
0, 1, inclusive="neither").any())
if test_input["threshold"]["Negatives"] != -1:
assert test_input["train_df"]["Negatives"][0] == 1
else:
assert test_input["train_df"]["Negatives"][0] == 0
def __init__(
self,
train_df_path=None,
test_df_path=None,
steps=None,
config_file=None,
params=None,
custom_reader=None,
):
steps = [
Parser().parse_dataset,
NullValuesHandler().execute,
Encoder().encode,
HandleOutlier().handle_outliers,
Scaler().execute,
SelectKBest().fit_transform,
Split().train_test_split,
]
super().__init__(
train_df_path=train_df_path,
test_df_path=test_df_path,
steps=steps,
config_file=config_file,
params=params,
custom_reader=custom_reader,
)
def test_incorrect_input_type(test_input, error):
with pytest.raises(error):
scaler = Scaler()
scaler.execute(params=test_input)
def test_MinMaxScaler_output(test_input):
scaler = Scaler()
scaler.execute(params=test_input)
assert test_input["df"]["Distance"].values.all() >= 0
assert test_input["df"]["Distance"].values.all() <= 1