def test_structured_data_input_transform():
(x, _), _1 = common.dataframe_dataframe()
input_node = node.StructuredDataInput()
input_node.fit(x)
input_node.transform(x)
assert input_node.column_names[0] == 'sex'
assert input_node.column_types == common.COLUMN_TYPES_FROM_CSV
def test_structured_data_input_name_type_mismatch():
(x, _), _1 = common.dataframe_dataframe()
with pytest.raises(ValueError) as info:
input_node = node.StructuredDataInput(
column_types=common.COLUMN_TYPES_FROM_CSV)
input_node.fit(x)
assert 'Column_names and column_types are mismatched.' in str(info.value)
def test_structured_data_input_name_type_mismatch():
(x, _), _1 = common.dataframe_dataframe()
column_types = copy.copy(common.COLUMN_TYPES_FROM_CSV)
column_types['age_'] = column_types.pop('age')
with pytest.raises(ValueError) as info:
input_node = node.StructuredDataInput(column_types=column_types)
input_node.transform(x)
assert 'Column_names and column_types are mismatched.' in str(info.value)
def test_structured_data_from_dataframe_dataframe_classifier(tmp_dir):
clf = ak.StructuredDataClassifier(directory=tmp_dir, max_trials=1)
(x, y), (val_x, val_y) = common.dataframe_dataframe()
clf.fit(x=x, y=y, epochs=2, validation_data=(val_x, val_y))
assert clf.predict(val_x).shape == (len(val_x), 1)