def it_can_apply_replace_strings(self, request, columns, derived_columns,
expected_new_columns, expected_inplace):
df = DataFrameMock.df_generic(sample_size=100)
get_df_from_csv_ = function_mock(request,
"trousse.dataset.get_df_from_csv")
get_df_from_csv_.return_value = df
dataset = Dataset(data_file="fake/path0")
pd_replace_ = method_mock(request, pd.Series, "replace")
pd_replace_.return_value = pd.Series([0] * 100)
replace_strings = replace_strings = fop.ReplaceStrings(
columns=columns,
derived_columns=derived_columns,
replacement_map={"a": "b"})
replaced_dataset = replace_strings._apply(dataset)
assert replaced_dataset is not None
assert replaced_dataset is not dataset
assert isinstance(replaced_dataset, Dataset)
for col in expected_new_columns:
assert col in replaced_dataset.data.columns
get_df_from_csv_.assert_called_once_with("fake/path0")
assert len(pd_replace_.call_args_list) == len(columns)
pd.testing.assert_series_equal(pd_replace_.call_args_list[0][0][0],
df[columns[0]])
assert pd_replace_.call_args_list[0][1] == {
"inplace": expected_inplace,
"to_replace": {
"a": "b"
},
}
def test_replace_strings(csv, columns, derived_columns, expected_csv):
dataset = Dataset(data_file=csv)
expected_df = load_expectation(expected_csv, type_="csv")
replace_strings = fop.ReplaceStrings(columns=columns,
derived_columns=derived_columns,
replacement_map={"d": "a"})
replaced_dataset = replace_strings(dataset)
pd.testing.assert_frame_equal(replaced_dataset.data, expected_df)
def it_construct_from_args(self, request):
_init_ = initializer_mock(request, fop.ReplaceStrings)
replace_strings = fop.ReplaceStrings(columns=["col0"],
derived_columns=["col1"],
replacement_map={"a": "b"})
_init_.assert_called_once_with(ANY,
columns=["col0"],
derived_columns=["col1"],
replacement_map={"a": "b"})
assert isinstance(replace_strings, fop.ReplaceStrings)
def it_knows_if_equal(self, other, expected_equal):
feat_op = fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"a": "b",
"c": "d"
},
)
equal = feat_op == other
assert type(equal) == bool
assert equal == expected_equal
def it_knows_its_str(self):
feat_op = fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"a": "b",
"c": "d"
},
)
_str = str(feat_op)
assert type(_str) == str
assert _str == (
"ReplaceStrings(\n\tcolumns=['exam_num_col_0'],\n\treplacement_map="
"{'a': 'b', 'c': 'd'},\n\tderived_columns=['replaced_exam_num_col_0'],\n)"
)
def and_it_validates_its_arguments(self, request):
validate_columns_ = method_mock(request, fop.ReplaceStrings,
"_validate_single_element_columns")
validate_derived_columns_ = method_mock(
request, fop.ReplaceStrings,
"_validate_single_element_derived_columns")
validate_replacement_map = method_mock(request, fop.ReplaceStrings,
"_validate_replacement_map")
replace_strings = fop.ReplaceStrings(columns=["col0"],
derived_columns=["col1"],
replacement_map={"a": "b"})
validate_columns_.assert_called_once_with(replace_strings, ["col0"])
validate_derived_columns_.assert_called_once_with(
replace_strings, ["col1"])
validate_replacement_map.assert_called_once_with(
replace_strings, {"a": "b"})
def it_can_replace_with_template_call(self, request):
_apply_ = method_mock(request, fop.ReplaceStrings, "_apply")
track_history_ = method_mock(request, Dataset, "track_history")
df = DataFrameMock.df_generic(sample_size=100)
get_df_from_csv_ = function_mock(request,
"trousse.dataset.get_df_from_csv")
get_df_from_csv_.return_value = df
dataset_in = Dataset(data_file="fake/path0")
dataset_out = Dataset(data_file="fake/path0")
_apply_.return_value = dataset_out
replace_strings = fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={"a": "b"},
)
replaced_dataset = replace_strings(dataset_in)
_apply_.assert_called_once_with(replace_strings, dataset_in)
track_history_.assert_called_once_with(replaced_dataset,
replace_strings)
assert replaced_dataset is dataset_out
def replacestrings_exam_num_col_0_replaced_exam_num_col_0_a_b():
return fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={"a": "b"},
)
class DescribeReplaceStrings:
def it_construct_from_args(self, request):
_init_ = initializer_mock(request, fop.ReplaceStrings)
replace_strings = fop.ReplaceStrings(columns=["col0"],
derived_columns=["col1"],
replacement_map={"a": "b"})
_init_.assert_called_once_with(ANY,
columns=["col0"],
derived_columns=["col1"],
replacement_map={"a": "b"})
assert isinstance(replace_strings, fop.ReplaceStrings)
def and_it_validates_its_arguments(self, request):
validate_columns_ = method_mock(request, fop.ReplaceStrings,
"_validate_single_element_columns")
validate_derived_columns_ = method_mock(
request, fop.ReplaceStrings,
"_validate_single_element_derived_columns")
validate_replacement_map = method_mock(request, fop.ReplaceStrings,
"_validate_replacement_map")
replace_strings = fop.ReplaceStrings(columns=["col0"],
derived_columns=["col1"],
replacement_map={"a": "b"})
validate_columns_.assert_called_once_with(replace_strings, ["col0"])
validate_derived_columns_.assert_called_once_with(
replace_strings, ["col1"])
validate_replacement_map.assert_called_once_with(
replace_strings, {"a": "b"})
@pytest.mark.parametrize(
"columns, derived_columns, expected_new_columns, expected_inplace",
[
(["exam_num_col_0"], ["col1"], ["col1"], False),
(["exam_num_col_0"], None, [], True),
],
)
def it_can_apply_replace_strings(self, request, columns, derived_columns,
expected_new_columns, expected_inplace):
df = DataFrameMock.df_generic(sample_size=100)
get_df_from_csv_ = function_mock(request,
"trousse.dataset.get_df_from_csv")
get_df_from_csv_.return_value = df
dataset = Dataset(data_file="fake/path0")
pd_replace_ = method_mock(request, pd.Series, "replace")
pd_replace_.return_value = pd.Series([0] * 100)
replace_strings = replace_strings = fop.ReplaceStrings(
columns=columns,
derived_columns=derived_columns,
replacement_map={"a": "b"})
replaced_dataset = replace_strings._apply(dataset)
assert replaced_dataset is not None
assert replaced_dataset is not dataset
assert isinstance(replaced_dataset, Dataset)
for col in expected_new_columns:
assert col in replaced_dataset.data.columns
get_df_from_csv_.assert_called_once_with("fake/path0")
assert len(pd_replace_.call_args_list) == len(columns)
pd.testing.assert_series_equal(pd_replace_.call_args_list[0][0][0],
df[columns[0]])
assert pd_replace_.call_args_list[0][1] == {
"inplace": expected_inplace,
"to_replace": {
"a": "b"
},
}
def it_can_replace_with_template_call(self, request):
_apply_ = method_mock(request, fop.ReplaceStrings, "_apply")
track_history_ = method_mock(request, Dataset, "track_history")
df = DataFrameMock.df_generic(sample_size=100)
get_df_from_csv_ = function_mock(request,
"trousse.dataset.get_df_from_csv")
get_df_from_csv_.return_value = df
dataset_in = Dataset(data_file="fake/path0")
dataset_out = Dataset(data_file="fake/path0")
_apply_.return_value = dataset_out
replace_strings = fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={"a": "b"},
)
replaced_dataset = replace_strings(dataset_in)
_apply_.assert_called_once_with(replace_strings, dataset_in)
track_history_.assert_called_once_with(replaced_dataset,
replace_strings)
assert replaced_dataset is dataset_out
@pytest.mark.parametrize(
"other, expected_equal",
[
(
fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"a": "b",
"c": "d"
},
),
True,
),
(
fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"c": "d",
"a": "b"
},
),
True,
),
(
fop.ReplaceStrings(
columns=["exam_num_col_1"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"a": "b",
"c": "d"
},
),
False,
),
(
fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_1"],
replacement_map={
"a": "b",
"c": "d"
},
),
False,
),
(
fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"a": "b",
},
),
False,
),
(
fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"c": "b",
},
),
False,
),
(dict(), False),
],
)
def it_knows_if_equal(self, other, expected_equal):
feat_op = fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"a": "b",
"c": "d"
},
)
equal = feat_op == other
assert type(equal) == bool
assert equal == expected_equal
def it_knows_its_str(self):
feat_op = fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={
"a": "b",
"c": "d"
},
)
_str = str(feat_op)
assert type(_str) == str
assert _str == (
"ReplaceStrings(\n\tcolumns=['exam_num_col_0'],\n\treplacement_map="
"{'a': 'b', 'c': 'd'},\n\tderived_columns=['replaced_exam_num_col_0'],\n)"
)
class DescribeTrousse:
@pytest.mark.parametrize(
"operations",
[
(),
(fop.FillNA(
columns=["nan"],
value=0,
), ),
(
fop.ReplaceStrings(
columns=["exam_num_col_0"],
derived_columns=["replaced_exam_num_col_0"],
replacement_map={"a": "b"},
),
fop.FillNA(columns=["nan"], value=0),
),
],
)
def it_contructs_from_args(self, request, operations):
_init_ = initializer_mock(request, fop.Trousse)
trousse = fop.Trousse(*operations)
_init_.assert_called_once_with(ANY, *operations)
assert isinstance(trousse, fop.Trousse)
def it_knows_its_operations(
self,
replacestrings_exam_num_col_0_replaced_exam_num_col_0_a_b,
fillna_col0_col1,
):
trousse = fop.Trousse(
replacestrings_exam_num_col_0_replaced_exam_num_col_0_a_b,
fillna_col0_col1)
operations = trousse.operations
assert type(operations) == tuple
assert operations == (
replacestrings_exam_num_col_0_replaced_exam_num_col_0_a_b,
fillna_col0_col1,
)
def it_knows_how_to_call(
self,
request,
replacestrings_exam_num_col_0_replaced_exam_num_col_0_a_b,
fillna_col0_col1,
):
dataset_in = instance_mock(request, Dataset, "in")
dataset_out_1 = instance_mock(request, Dataset, "1")
dataset_out_2 = instance_mock(request, Dataset, "2")
_call_replacestrings = method_mock(request, fop.ReplaceStrings,
"__call__")
_call_replacestrings.return_value = dataset_out_1
_call_fillna = method_mock(request, fop.FillNA, "__call__")
_call_fillna.return_value = dataset_out_2
trousse = fop.Trousse(
replacestrings_exam_num_col_0_replaced_exam_num_col_0_a_b,
fillna_col0_col1)
new_dataset = trousse(dataset_in)
_call_replacestrings.assert_called_once_with(
replacestrings_exam_num_col_0_replaced_exam_num_col_0_a_b,
dataset_in)
_call_fillna.assert_called_once_with(fillna_col0_col1, dataset_out_1)
assert isinstance(new_dataset, Dataset)
assert new_dataset == dataset_out_2
def it_knows_its_str(self, fillna_col0_col1, fillna_col1_col4):
trousse = fop.Trousse(fillna_col0_col1, fillna_col1_col4)
_str = str(trousse)
assert type(_str) == str
assert _str == (
"Trousse: (FillNA(\n\tcolumns=['col0'],\n\tvalue=0,\n\t"
"derived_columns=['col1'],\n), FillNA(\n\tcolumns=['col1']"
",\n\tvalue=0,\n\tderived_columns=['col4'],\n))")