def test_should_return_same_df_when_column_not_exists(self):
transformation = Transformation(self.test_data)
transformation.drop(["xpto"])
transformed_df = transformation.dataframe.collect()
self.assertEqual(transformed_df, self.test_data.collect())
def test_should_have_same_df_when_column_list_is_empty(self):
transformation = Transformation(self.test_data)
transformation.drop([])
transformed_df = transformation.dataframe.collect()
self.assertEqual(transformed_df, self.test_data.collect())
def test_should_remove_one_column_from_dataframe(self):
transformation = Transformation(self.test_data)
transformation.drop(["alert"])
current_result = transformation.dataframe.columns
expected_result = self.spark.createDataFrame([
(1704567252, "California", 0.82, "Automatic",
[-116.8, 33.3333333, 12.04]),
(1391707828, "Alaska", 1.1, None, [-148.942, 64.9081, 10.6]),
(1435498694, "Chile", 4.9, "Reviewed", [-70.6202, -21.4265, 52.24
]),
(1609879110, "Hawaii", 2.0099, "Automatic",
[-155.429000854492, 19.2180004119873, 33.2999992370605]),
(1224994646, "Indonesia", 4.8, "Reviewed", [126.419, 0.2661, 10]),
(1801059964, "Nevada", 0.5, "Automatic", [-116.242, 36.7564, 0.8]),
(1262739669, "Arkansas", 1.9, "Reviewed",
[-91.4295, 35.863, 16.41]),
(1890118874, "Montana", 1.33, "Reviewed",
[-110.434, 44.4718333, 2.21]),
(1025727100, "Oklahoma", 1.58, "Reviewed",
[-98.53233333, 36.57083333, 6.31]),
(1834567116, "Idaho", 2.6, "Reviewed", [-115.186, 44.2666, 10])
], ["date", "place", "mag", "status", "coordinates"]).columns
self.assertEqual(current_result, expected_result)
def test_should_remove_two_columns_from_dataframe(self):
transformation = Transformation(self.test_data)
transformation.drop(["coordinates", "alert"])
current_result = transformation.dataframe.columns
expected_result = self.spark.createDataFrame(
[(1704567252, "California", 0.82, "Automatic"),
(1391707828, "Alaska", 1.1, None),
(1435498694, "Chile", 4.9, "Reviewed"),
(1609879110, "Hawaii", 2.0099, "Automatic"),
(1224994646, "Indonesia", 4.8, "Reviewed"),
(1801059964, "Nevada", 0.5, "Automatic"),
(1262739669, "Arkansas", 1.9, "Reviewed"),
(1890118874, "Montana", 1.33, "Reviewed"),
(1025727100, "Oklahoma", 1.58, "Reviewed"),
(1834567116, "Idaho", 2.6, "Reviewed")],
["date", "place", "mag", "status"]).columns
self.assertEqual(current_result, expected_result)
def test_should_return_transformed_data_using_all_pipeline_components(
self, mock_get_data):
self.create_tmp_folder()
fake_api_input = ApiInput(self.FAKE_URL)
mock_get_data.return_value = self.FAKE_INPUT_DATA
extraction_process = Extraction(fake_api_input)
extraction_process.extract()
raw_data = extraction_process.data
raw_df = self.spark.createDataFrame(
raw_data,
["date", "place", "mag", "status", "coordinates", "alert"])
transformation_process = Transformation(raw_df)
transformation_process.drop(["alert"])
transformation_process.rename({"mag": "magnitude", "place": "city"})
transformation_process.replace_null_values({"status": "Automatic"})
transformation_process.lowercase(["status"])
transformation_process.convert_data_type({"date": IntegerType()})
transformation_process.split_content(
"coordinates", ["longitude", "latitude", "depth"])
transformed_df = transformation_process.dataframe
csv_storage = CsvStorage(self.OUTPUT_FILEPATH)
loading_process = Loading(csv_storage)
loading_process.load(transformed_df)
current_result = self.spark \
.read \
.csv(self.OUTPUT_FILEPATH, header=True, inferSchema=True) \
.collect()
expected_result = self.spark \
.createDataFrame(self.FAKE_EXPECTED_DATA,
["date", "city", "magnitude", "status", "longitude", "latitude", "depth"]) \
.collect()
self.assertEqual(current_result, expected_result)
self.delete_test_file()