def test_transform_drops_unnecessary_columns(self):
ccn = preprocessing.CheckColumnNames()
input_data = pd.DataFrame({
"one": [1, 2],
"two": [3, 4],
"three": [5, 6]
})
test_data = pd.DataFrame({
"one": [7, 8],
"two": [9, 10],
"three": [11, 12],
"four": [13, 14]
})
expected_data = pd.DataFrame({
"one": [7, 8],
"two": [9, 10],
"three": [11, 12]
})
#Ensure columns are in a particular order:
input_data = input_data[["one", "two", "three"]]
expected_data = expected_data[["one", "two", "three"]]
ccn.fit(input_data)
pd.util.testing.assert_frame_equal(ccn.transform(test_data),
expected_data)
def test_transform_data_has_wrong_columns(self):
ccn = preprocessing.CheckColumnNames()
input_data = pd.DataFrame({"one": [1, 2], "two": [3, 4]})
ccn.fit(input_data)
test_data = pd.DataFrame({"one": [1, 2], "three": [3, 4]})
with pytest.raises(KeyError):
ccn.transform(test_data)
def test_transform_with_user_specified_colums(self):
ccn = preprocessing.CheckColumnNames(column_names=["c", "b", "a"])
input_data = pd.DataFrame({"e": [-2, -1, 0],
"a": [1, 2, 3],
"b": [4, 5, 6],
"c": [7, 8, 9],
"d": [10, 11, 12]})
expected_data = pd.DataFrame({"c": [7, 8, 9],
"b": [4, 5, 6],
"a": [1, 2, 3]})
expected_data = expected_data[["c", "b", "a"]]
transformed_data = ccn.transform(input_data)
pd.util.testing.assert_frame_equal(expected_data, transformed_data)
def test_transform_reorders_columns(self):
ccn = preprocessing.CheckColumnNames()
input_data = pd.DataFrame({"one": [1, 2],
"two": [3, 4],
"three": [5, 6]})
test_data = pd.DataFrame({"one": [7, 8],
"two": [9, 10],
"three": [11, 12]})
expected_data = test_data.copy()
#Ensure columns are in a particular order:
input_data = input_data[["one", "two", "three"]]
test_data = test_data[["two", "one", "three"]]
expected_data = expected_data[["one", "two", "three"]]
with pytest.raises(AssertionError):
pd.util.testing.assert_frame_equal(test_data, expected_data)
ccn.fit(input_data)
pd.util.testing.assert_frame_equal(ccn.transform(test_data), expected_data)
def test_transform_called_before_fit(self):
ccn = preprocessing.CheckColumnNames()
data = pd.DataFrame()
with pytest.raises(NotFittedError):
ccn.transform(data)