def test_index_date():
expected_indx_dates = {
19: datetime.datetime(3397, 8, 4, 0, 0),
99: datetime.datetime(2727, 3, 17, 0, 0),
24581: datetime.datetime(3401, 6, 23, 0, 0),
3014: datetime.datetime(2581, 5, 1, 0, 0)
}
calculate_index_date(events_df, mortality_df, deliverables_path)
indx_date_df = pd.read_csv(deliverables_path + 'etl_index_dates.csv')
indx_date_df['indx_date'] = indx_date_df['indx_date'].apply(date_convert)
indx_date = dict(zip(indx_date_df.patient_id, indx_date_df.indx_date))
if isinstance(indx_date, pd.DataFrame):
indx_date = dict(zip(indx_date.patient_id, indx_date.indx_date))
res = True
if len(indx_date) != len(expected_indx_dates):
res = False
else:
for key, value in expected_indx_dates.iteritems():
if key not in indx_date:
res = False
break
if not abs(indx_date[key].date() == value.date()):
res = False
break
eq_(res, True, "Index dates do not match")
def test_index_date():
expected_indx_dates = {
19: datetime.datetime(3397, 8, 4, 0, 0),
99: datetime.datetime(2727, 3, 17, 0, 0),
24581: datetime.datetime(3401, 6, 23, 0, 0),
3014: datetime.datetime(2581, 5, 1, 0, 0),
}
calculate_index_date(events_df, mortality_df, deliverables_path)
indx_date_df = pd.read_csv(deliverables_path + "etl_index_dates.csv")
indx_date_df["indx_date"] = indx_date_df["indx_date"].apply(date_convert)
indx_date = dict(zip(indx_date_df.patient_id, indx_date_df.indx_date))
if isinstance(indx_date, pd.DataFrame):
indx_date = dict(zip(indx_date.patient_id, indx_date.indx_date))
res = True
if len(indx_date) != len(expected_indx_dates):
res = False
else:
for key, value in expected_indx_dates.iteritems():
if key not in indx_date:
res = False
break
if not abs(indx_date[key].date() == value.date()):
res = False
break
eq_(res, True, "Index dates do not match")
def test_aggregate_events():
expected_data = []
with open('tests/expected_etl_aggregated_events.csv') as expected_file:
expected_data = expected_file.readlines()
expected_data = expected_data[1:]
expected_data.sort()
indx_dates_df = calculate_index_date(events_df, mortality_df,
deliverables_path)
filtered_events_df = filter_events(events_df, indx_dates_df,
deliverables_path)
aggregated_events_df = aggregate_events(filtered_events_df, mortality_df,
feature_map_df, deliverables_path)
actual_data = []
with open(deliverables_path + 'etl_aggregated_events.csv') as actual_file:
actual_data = actual_file.readlines()
actual_data = actual_data[1:]
actual_data.sort()
res = True
msg = ""
for idx, line in enumerate(expected_data):
first = line.split(',')
second = actual_data[idx].split(',')
if not (float(first[0]) == float(second[0])
and float(first[1]) == float(second[1])
and abs(float(first[2]) - float(second[2])) <= 0.1):
res = False
msg = "Mistmatch on line %d. \n\nExpected: %s \nActual: %s " % (
idx + 1, line, actual_data[idx])
break
eq_(res, True, "Aggregated events do not match. " + msg)
def test_aggregate_events():
expected_data = []
with open("tests/expected_etl_aggregated_events.csv") as expected_file:
expected_data = expected_file.readlines()
expected_data = expected_data[1:]
expected_data.sort()
indx_dates_df = calculate_index_date(events_df, mortality_df, deliverables_path)
filtered_events_df = filter_events(events_df, indx_dates_df, deliverables_path)
aggregated_events_df = aggregate_events(filtered_events_df, mortality_df, feature_map_df, deliverables_path)
actual_data = []
with open(deliverables_path + "etl_aggregated_events.csv") as actual_file:
actual_data = actual_file.readlines()
actual_data = actual_data[1:]
actual_data.sort()
res = True
msg = ""
for idx, line in enumerate(expected_data):
first = line.split(",")
second = actual_data[idx].split(",")
if not (
float(first[0]) == float(second[0])
and float(first[1]) == float(second[1])
and abs(float(first[2]) - float(second[2])) <= 0.1
):
res = False
msg = "Mistmatch on line %d. \n\nExpected: %s \nActual: %s " % (idx + 1, line, actual_data[idx])
break
eq_(res, True, "Aggregated events do not match. " + msg)