def test_target_imputer_woodwork_custom_overrides_returned_by_components(
y_pd, has_nan, impute_strategy):
y_to_use = y_pd.copy()
if has_nan:
y_to_use[len(y_pd) - 1] = np.nan
override_types = [Integer, Double, Categorical, Boolean]
for logical_type in override_types:
try:
y = ww.DataColumn(y_to_use.copy(), logical_type=logical_type)
except TypeError:
continue
impute_strategy_to_use = impute_strategy
if logical_type in [Categorical, NaturalLanguage]:
impute_strategy_to_use = "most_frequent"
imputer = TargetImputer(impute_strategy=impute_strategy_to_use)
imputer.fit(None, y)
_, y_t = imputer.transform(None, y)
assert isinstance(y_t, ww.DataColumn)
if impute_strategy_to_use == "most_frequent" or not has_nan:
assert y_t.logical_type == logical_type
else:
assert y_t.logical_type == Double
def test_target_imputer_all_bool_return_original(data_type, make_data_type):
y = pd.Series([True, True, False, True, True], dtype=bool)
y = make_data_type(data_type, y)
y_expected = pd.Series([True, True, False, True, True], dtype='boolean')
imputer = TargetImputer()
imputer.fit(None, y)
_, y_t = imputer.transform(None, y)
assert_series_equal(y_expected, y_t.to_series())
def test_target_imputer_no_y(X_y_binary):
X, y = X_y_binary
imputer = TargetImputer()
assert imputer.fit_transform(None, None) == (None, None)
imputer = TargetImputer()
imputer.fit(None, None)
assert imputer.transform(None, None) == (None, None)
def test_target_imputer_boolean_dtype(data_type, make_data_type):
y = pd.Series([True, np.nan, False, np.nan, True], dtype='boolean')
y_expected = pd.Series([True, True, False, True, True], dtype='boolean')
y = make_data_type(data_type, y)
imputer = TargetImputer()
imputer.fit(None, y)
_, y_t = imputer.transform(None, y)
assert_series_equal(y_expected, y_t.to_series())
def test_target_imputer_fit_transform_all_nan_empty():
y = pd.Series([np.nan, np.nan])
imputer = TargetImputer()
imputer.fit(None, y)
with pytest.raises(RuntimeError, match="Transformed data is empty"):
imputer.transform(None, y)
imputer = TargetImputer()
with pytest.raises(RuntimeError, match="Transformed data is empty"):
imputer.fit_transform(None, y)
def test_target_imputer_col_with_non_numeric_with_numeric_strategy():
y = pd.Series([np.nan, "a", "b"])
imputer = TargetImputer(impute_strategy='mean')
with pytest.raises(ValueError,
match="Cannot use mean strategy with non-numeric data"):
imputer.fit_transform(None, y)
with pytest.raises(ValueError,
match="Cannot use mean strategy with non-numeric data"):
imputer.fit(None, y)
imputer = TargetImputer(impute_strategy='median')
with pytest.raises(
ValueError,
match="Cannot use median strategy with non-numeric data"):
imputer.fit_transform(None, y)
with pytest.raises(
ValueError,
match="Cannot use median strategy with non-numeric data"):
imputer.fit(None, y)
def test_target_imputer_does_not_reset_index():
y = pd.Series(np.arange(10))
y[5] = np.nan
assert y.index.tolist() == list(range(10))
y.drop(0, inplace=True)
pd.testing.assert_series_equal(
pd.Series([1, 2, 3, 4, np.nan, 6, 7, 8, 9],
dtype=float,
index=list(range(1, 10))), y)
imputer = TargetImputer(impute_strategy="mean")
imputer.fit(None, y=y)
_, y_t = imputer.transform(None, y)
pd.testing.assert_series_equal(
pd.Series([1.0, 2, 3, 4, 5, 6, 7, 8, 9],
dtype=float,
index=list(range(1, 10))), y_t.to_series())