def test_multiple_args(self):
self.dataframe = pd.read_csv("datasets/encoding/testnew.csv")
with pytest.raises(Exception):
handler = NullValuesHandler()
handler.execute(
{"df": self.dataframe, "drop": True, "fill_missing": "mean"}
)
def test_categorical_replace(test_input2):
test_input2["train_df"] = pd.read_csv("datasets/encoding/test2.csv")
test_input2["test_df"] = pd.read_csv("datasets/encoding/test2.csv")
handler = NullValuesHandler()
handler.execute(params=test_input2)
assert (len(test_input2["train_df"][test_input2["train_df"].Capitals ==
"xyz"]["Capitals"]) == 8)
assert (len(test_input2["test_df"][test_input2["test_df"].Capitals ==
"xyz"]["Capitals"]) == 8)
def test_categorical_drops(test_input1):
train_row_count = (test_input1["train_df"].shape)[0]
test_row_count = (test_input1["test_df"].shape)[0]
handler = NullValuesHandler()
handler.execute(params=test_input1)
assert "0" not in test_input1["train_df"]["Other Capitals"]
assert (test_input1["train_df"].shape)[0] <= train_row_count
assert "0" not in test_input1["test_df"]["Other Capitals"]
assert (test_input1["test_df"].shape)[0] <= test_row_count
def test_drop_col():
params = {
"train_df": dataframe1,
"test_df": dataframe1,
"drop": True,
"column_list": ["Distance"],
}
handler = NullValuesHandler()
handler.execute(params=params)
assert "Distance" not in params["train_df"].columns
assert "Distance" not in params["test_df"].columns
def test_auto():
dataframe1 = pd.read_csv("datasets/encoding/test2.csv")
params = {"train_df": dataframe1, "test_df": dataframe1}
r = (params["train_df"].shape)[0]
rt = (params["test_df"].shape)[0]
handler = NullValuesHandler()
handler.execute(params=params)
r_new = (params["train_df"].shape)[0]
rt_new = (params["test_df"].shape)[0]
assert r_new != r
assert rt_new != rt
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_output(test_input3):
handler = NullValuesHandler()
handler.execute(params=test_input3)
assert not test_input3["train_df"].isnull().any()["Distance"]
assert not test_input3["test_df"].isnull().any()["Distance"]
def test_incorrect_input_type(error, test_input):
with pytest.raises(error):
handler = NullValuesHandler()
handler.execute(params=test_input)
def test_null_dataframe():
with pytest.raises(ValueError):
handler = NullValuesHandler()
handler.execute({})
def test_incorrect_input_type(self):
# for dataframe argument
array = np.random.random((5, 5))
with pytest.raises(TypeError):
handler = NullValuesHandler()
handler.execute({"df": array})
# for drop argument
self.dataframe = pd.read_csv("datasets/encoding/testnew.csv")
with pytest.raises(TypeError):
handler = NullValuesHandler()
handler.execute({"df": self.dataframe, "drop": "nice"})
# for fill_missing argument
with pytest.raises(TypeError):
handler = NullValuesHandler()
handler.execute({"df": dataframe, "fill_missing": 3})
with pytest.raises(ArgumentsError):
handler = NullValuesHandler()
handler.execute({"df": dataframe, "fill_missing": "sum"})
# for fill_values argument
with pytest.raises(TypeError):
value = [5]
handler = NullValuesHandler()
handler.execute({"df": dataframe, "fill_values": value})
with pytest.raises(ArgumentsError):
value = {"Label": 10}
handler = NullValuesHandler()
handler.execute({"df": dataframe, "fill_values": value})
def test_multiple_args(self):
with pytest.raises(ArgumentsError):
handler = NullValuesHandler()
handler.execute({
"df": dataframe,
"drop": True,
"fill_missing": "mean"
})
with pytest.raises(ArgumentsError):
handler = NullValuesHandler()
handler.execute({
"df": dataframe,
"drop": True,
"fill_values": {
"Test": "Tata"
},
})
with pytest.raises(ArgumentsError):
handler = NullValuesHandler()
handler.execute({
"df": dataframe,
"drop": True,
"fill_values": {
"Test": "Tata"
},
})
with pytest.raises(ArgumentsError):
handler = NullValuesHandler()
handler.execute({
"df": dataframe,
"fill_missing": "mean",
"fill_values": {
"Test": "Tata"
},
})
def test_none_args(self):
with pytest.raises(ArgumentsError):
handler = NullValuesHandler()
handler.execute({"df": dataframe})
def test_drop_col(self):
params = {"df": dataframe, "drop": True, "column_list": ["Distance"]}
handler = NullValuesHandler()
handler.execute(params=params)
assert "Distance" not in params["df"].columns
def test_incorrect_input_type(self):
self.dataframe = pd.read_csv("datasets/encoding/testnew.csv")
with pytest.raises(TypeError):
handler = NullValuesHandler()
handler.execute({"df": self.dataframe, "drop": "nice"})
def test_output(self):
params = {"df": dataframe, "drop": True}
handler = NullValuesHandler()
handler.execute(params=params)
assert not params["df"].isnull().values.any()
params = {"df": dataframe, "fill_missing": "mean"}
handler = NullValuesHandler()
handler.execute(params=params)
assert not params["df"].isnull().any()["Distance"]
params = {"df": dataframe, "fill_values": {"Distance": 0}}
handler = NullValuesHandler()
handler.execute(params=params)
assert not params["df"].isnull().any()["Distance"]
def test_multiple_args(test_input):
with pytest.raises(ArgumentsError):
handler = NullValuesHandler()
handler.execute(params=test_input)
def test_output(test_input1):
handler = NullValuesHandler()
handler.execute(params=test_input1)
assert not test_input1["df"].isnull().any()["Distance"]