def test_extract_time_periods_by_dates_invalid_test_data():
test_data = pd.DataFrame(create_test_data())
# this is invalid test data as there is no 'end_date' column
invalid_test_data = test_data.drop('end_date', axis=1)
initial_date = '2016-02-02'
final_date = '2017-02-01'
with pytest.raises(KeyError):
get_transform_data.extract_time_periods_by_dates(
invalid_test_data, initial_date, final_date)
def test_extract_time_periods_by_dates_invalid_test_dates():
test_data = pd.DataFrame(create_test_data())
test_data.loc[:, 'end_date'] = pd.to_datetime(test_data['end_date'])
# this is an invalid date because the initial and final dates are not within 1 year
invalid_initial_date = '2016-02-02'
invalid_final_date = '2017-02-02'
with pytest.raises(ValueError):
get_transform_data.extract_time_periods_by_dates(
test_data, invalid_initial_date, invalid_final_date)
def time_series(client, data_series, initial_date, final_date):
"""
retrieves a sub-time series of a time series associated with a gro data_series
:param client: GroClient
:param data_series: A dictionary of Gro data_series
:param final_date: 'YYYY-MM-DD'
:param initial_date: 'YYYY-MM-DD'
:return: A dataframe with data from the relevant dates
"""
logger = client.get_logger()
entities = {'metric_id', 'item_id', 'region_id', 'source_id', 'frequency_id', 'start_date'}
discard = []
for entity_type in data_series:
if entity_type not in entities:
discard.append(entity_type)
for entity_type in discard:
data_series.pop(entity_type)
data = get_transform_data.get_data(client, **data_series)
consolidated_data = get_transform_data.combine_subregions(data)
try:
ts = get_transform_data.extract_time_periods_by_dates(consolidated_data,
initial_date,
final_date)
return ts
except Exception as e:
message = ('Please check availability of data for {}'.format(client.lookup(
'items', data_series['item_id'])['name']))
logger.warning(message)
raise e
def test_extract_time_periods_by_dates_with_non_unique_dates():
test_data = pd.DataFrame(create_test_data())
initial_date = '2019-03-01'
final_date = '2019-07-31'
expected_data = test_data
expected_data.loc[:, 'date'] = pd.to_datetime(expected_data['end_date'])
expected_data.loc[:, 'year'] = expected_data['date'].dt.year.astype('str')
expected_data.loc[:, 'period'] = ['2019-03-01 to 2019-07-31',
'2005-03-01 to 2005-07-31',
'2000-03-01 to 2000-07-31',
'2005-03-01 to 2005-07-31']
expected_data = expected_data[expected_data.period != '2005-03-01 to 2005-07-31']
expected_data.loc[:, 'mm-dd'] = expected_data['date'].dt.strftime("%m-%d")
expected_data = expected_data.sort_values(by=list(expected_data.columns), axis=0,
ascending=False)
assert_frame_equal(
get_transform_data.extract_time_periods_by_dates(test_data,
initial_date,
final_date),
expected_data)
