def test_get_column_indices_pandas_nonunique_columns_error(key):
pd = pytest.importorskip("pandas")
toy = np.zeros((1, 5), dtype=int)
columns = ["col1", "col1", "col2", "col3", "col2"]
X = pd.DataFrame(toy, columns=columns)
err_msg = "Selected columns, {}, are not unique in dataframe".format(key)
with pytest.raises(ValueError) as exc_info:
_get_column_indices(X, key)
assert str(exc_info.value) == err_msg
def _grid(self, X, features, percentiles, grid_resolution):
features_indices = np.asarray(
_get_column_indices(X, features), dtype=np.int32, order='C'
).ravel()
grid, values = _grid_from_X(
_safe_indexing(X, features_indices, axis=1), percentiles,
grid_resolution
)
return grid, values
def accumulated_local_effects(est, x, feature, n_quantiles):
"""calculates ale for a feature"""
ale = np.array(n_quantiles,)
features_indices = np.asarray(
_get_column_indices(x, feature), dtype=np.int32, order='C'
).ravel()
quantiles = _quantiles_from_x(_safe_indexing(x, features_indices, axis=1), n_quantiles)
x_feat = _safe_indexing(x, feature, axis=1)
if x_feat.to_numpy().dtype.name == "category" or x_feat.to_numpy().dtype == "object":
ale = _ale_for_categorical(est, quantiles, x, x_feat)
else:
ale = _ale_for_numeric(est, quantiles, x, x_feat)
return ale
def test_get_column_indices_error(key, err_msg):
pd = pytest.importorskip("pandas")
X_df = pd.DataFrame(X_toy, columns=['col_0', 'col_1', 'col_2'])
with pytest.raises(ValueError, match=err_msg):
_get_column_indices(X_df, key)
