def setup(self):
"""Setup values for lines to claim tests."""
self.split_claim_columns = row_handling.csv_to_query_output(
SPLIT_CLAIM_CSV_PATH)[0]
self.split_claim_rows = row_handling.csv_to_query_output(
SPLIT_CLAIM_CSV_PATH)[1]
self.claim_reader = claim_reader.ClaimsDataReader()
def test_assert_split_claims_have_same_header_level_values():
"""The method should raise an error if two claims do not have the same header fields."""
reader = claim_reader.ClaimsDataReader()
sample_rows = test_helpers.fetch_sample_teradata_rows()
with pytest.raises(AssertionError):
reader._lines_to_claim(claim_lines=sample_rows,
columns=sample_rows[0].columns)
def __init__(
self,
start_date,
end_date,
measures,
infer_performance_period):
"""
Initialize Processor.
:param start_date: Processing start date
:type start_date: date
:param end_date: Processing end date
:type end_date: date
:param measures: Measures to calcuate
:type measures: list
"""
self.start_date = start_date
self.end_date = end_date
self.measures = measures
self.measure_calculators = measure_mapping.get_measure_calculators(measures)
self.measure_definitions = [
calculator.measure_definition for calculator in self.measure_calculators.values()
]
self.infer_performance_period = infer_performance_period
self.claim_reader = claim_reader.ClaimsDataReader()
self.session = teradata_connector.teradata_connection()
self.count = 0
self.count_no_claims = 0
self.count_errors = 0
def get_single_claim_with_quality_codes():
"""Load single claim list from test_single_claim.csv."""
reader = claim_reader.ClaimsDataReader()
claims = reader.load_from_csv(
csv_path='tests/assets/test_single_claim.csv',
provider_tin='tax_num',
provider_npi='npi_num')
return claims
def test_assert_split_claims_have_same_header_level_values_different_birth_date(
):
"""The method should NOT raise an error if two claims do not have the same birth date."""
reader = claim_reader.ClaimsDataReader()
columns, rows = row_handling.csv_to_query_output(SPLIT_CLAIM_CSV_PATH)
# Change the first birth date so that the two are different.
rows[0]['clm_ptnt_birth_dt'] = datetime.date(1000, 1, 1)
claim = reader._lines_to_claim(claim_lines=rows, columns=columns)
assert claim
def test_two_claim_data_reader(self):
"""Test the merging of claim lines into claims."""
reader = claim_reader.ClaimsDataReader()
claims = reader.load_from_csv(TWO_CLAIMS_CSV_PATH_FROM_CSV, 'tax_num',
'npi_num')
assert len(claims) == 2
for claim in claims:
assert len(claim.claim_lines) == 2
def test_single_claim_data_reader(self):
"""Test the parsing of a single claim into the Claim model object."""
reader = claim_reader.ClaimsDataReader()
claims = reader.load_from_csv(SINGLE_CLAIM_CSV_PATH, 'tax_num',
'npi_num')
assert len(claims) == 1
claim = claims[0]
assert len(claim.claim_lines) == 2
assert len(claim.dx_codes) == 12
assert len(claim.claim_lines[0].mdfr_cds) == 5
def test_query_returns_empty(self, mock_query_claims_from_teradata_batch):
"""Verify list of claims is empty if query returns no results."""
mock_query_claims_from_teradata_batch.return_value = row_handling.csv_to_query_output(
EMPTY_QUERY_CSV_PATH)
reader = claim_reader.ClaimsDataReader()
claims = reader.load_batch_from_db(provider_tin_list=['tax_num'],
provider_npi_list=['npi_num'],
start_date='start_date',
end_date='end_date')
assert len(claims) == 0
def test_batch_load_no_data(self, mock_config,
query_claims_from_teradata_batch_provider):
"""Test that when hide_sensitive_information is true, the rows are anonymized."""
mock_config.get.side_effect = mocking_config.config_side_effect(
{'hide_sensitive_information': False})
query_claims_from_teradata_batch_provider.return_value = ([], None)
reader = claim_reader.ClaimsDataReader()
output = reader.load_batch_from_db(['tax_num'], ['npi_num'],
datetime.date.today(),
datetime.date.today())
assert output == {}
def test_batch_load(self, mock_config,
query_claims_from_teradata_batch_provider):
"""Test that when hide_sensitive_information is true, the rows are anonymized."""
mock_config.get.side_effect = mocking_config.config_side_effect(
{'hide_sensitive_information': False})
query_claims_from_teradata_batch_provider.return_value = row_handling.csv_to_query_output(
TWO_CLAIMS_CSV_PATH)
reader = claim_reader.ClaimsDataReader()
output = reader.load_batch_from_db(['tax_num'], ['npi_num'],
datetime.date.today(),
datetime.date.today())
assert len(output[('tax_num', 'npi_num')]) == 2
def test_query_hides_sensitive_information_fewer_than_fifty(
self, mock_config, mock_query_claims_from_teradata_batch):
"""Test case when fewer than 50 claims and hide_sensitive_information is true."""
mock_config.get.side_effect = mocking_config.config_side_effect(
{'hide_sensitive_information': True})
provider_identifier = ('tax_num', 'npi_num')
sample_columns, sample_rows = row_handling.csv_to_query_output(
TWO_CLAIMS_CSV_PATH)
mock_query_claims_from_teradata_batch.return_value = (sample_columns,
sample_rows)
reader = claim_reader.ClaimsDataReader()
output = reader.load_batch_from_db(
['tax_num'], ['npi_num'], datetime.date.today(),
datetime.date.today())[provider_identifier]
assert output == []
def test_two_claim_data_reader(self, mock_query_claims_from_teradata_batch,
get):
"""Test the merging of claim lines into claims."""
mock_query_claims_from_teradata_batch.return_value = row_handling.csv_to_query_output(
TWO_CLAIMS_CSV_PATH)
get.side_effect = mocking_config.config_side_effect(
{'hide_sensitive_information': False})
provider_identifier = ('tax_num', 'npi_num')
reader = claim_reader.ClaimsDataReader()
claims = reader.load_batch_from_db(
provider_tin_list=['tax_num'],
provider_npi_list=['npi_num'],
start_date='start_date',
end_date='end_date')[provider_identifier]
assert len(claims) == 2
for claim in claims:
assert len(claim.claim_lines) == 2
def test_single_claim_data_reader(self,
mock_query_claims_from_teradata_batch,
get):
"""Test the parsing of a single claim into the Claim model object."""
mock_query_claims_from_teradata_batch.return_value = row_handling.csv_to_query_output(
SINGLE_CLAIM_CSV_PATH)
get.side_effect = mocking_config.config_side_effect(
{'hide_sensitive_information': False})
provider_identifier = ('tax_num', 'npi_num')
reader = claim_reader.ClaimsDataReader()
claims = reader.load_batch_from_db(
provider_tin_list=['tax_num'],
provider_npi_list=['npi_num'],
start_date='start_date',
end_date='end_date')[provider_identifier]
assert len(claims) == 1
claim = claims[0]
assert len(claim.claim_lines) == 2
assert len(claim.dx_codes) == 12
assert len(claim.claim_lines[0].mdfr_cds) == 5
def test_query_hides_sensitive_information_more_than_fifty(
self, mock_config, mock_query_claims_from_teradata_batch):
"""Test case when more than 50 claims and hide_sensitive_information is true."""
mock_config.get.side_effect = mocking_config.config_side_effect(
{'hide_sensitive_information': True})
provider_identifier = ('tax_num', 'npi_num')
sample_columns, sample_rows = row_handling.csv_to_query_output(
TWO_CLAIMS_CSV_PATH)
# Need > 50 rows for them to pass through anonymization.
sample_rows = sample_rows * 50
mock_query_claims_from_teradata_batch.return_value = (sample_columns,
sample_rows)
reader = claim_reader.ClaimsDataReader()
output = reader.load_batch_from_db(
['tax_num'], ['npi_num'], datetime.date.today(),
datetime.date.today())[provider_identifier]
# The same number of claims should be returned.
assert len(output) == 2 * 50 # Two claims in initial csv.
# The values of the fields in the returned rows should be different.
assert output[0]['clm_ptnt_birth_dt'] != sample_rows[0][
'clm_ptnt_birth_dt']
def _main(**kwargs):
"""Calculate measures for a given provider's data."""
if kwargs['measure'] in ['046', '407', '415', '416']:
raise NotImplementedError(
'Unable to calulate measure {} at this time. Calculation requires access to IDR.'
)
reader = claim_reader.ClaimsDataReader()
claims = reader.load_from_csv(kwargs['csv_input_path'],
provider_tin=kwargs['provider_tin'],
provider_npi=kwargs['provider_npi'])
measure = measure_mapping.get_measure_calculator(kwargs['measure'])
if all(key in kwargs for key in ['start_date', 'end_date']):
claims = claim_filtering.filter_claims_by_date(
claims_data=claims,
from_date=kwargs['start_date'],
to_date=kwargs['end_date'],
)
results = measure.execute(claims)
print(json.dumps(results, indent=4))
def test_filter_by_provider_tax(self):
"""Verify the claim is excluded if the provider npi does not match the one provided."""
reader = claim_reader.ClaimsDataReader()
claims = reader.load_from_csv(SINGLE_CLAIM_CSV_PATH,
'non_matching_tax_num', 'npi_num')
assert len(claims) == 0
def test_query_returns_empty(self):
"""Verify list of claims is empty if csv contains no results."""
reader = claim_reader.ClaimsDataReader()
claims = reader.load_from_csv(EMPTY_QUERY_CSV_PATH, 'tax_num',
'npi_num')
assert len(claims) == 0
