def test_big_data():
acs = ACS()
variables = ["B01001_002E"] * 50 + ['NAME']
geography = 'county'
inside = 'state:01'
result = acs.data(variables, geography, inside)
assert result.shape[1] == 4
assert set(result.columns) == {'B01001_002E', 'NAME', 'county', 'state'}
def test_big_query():
acs = ACS()
variables = ["B01001_002E"] * 50
geography = 'county'
inside = 'state:01'
result = acs._query(variables, geography, inside)
assert len(result) > 0
assert len(result[0]) == 52
def test_simple_query():
acs = ACS()
variables = ["NAME", "B01001_002E", "B01001_026E"]
geography = 'county'
inside = 'state:01'
result = acs._query(variables, geography, inside)
assert len(result) > 0
assert len(result[0]) == 5
def test_hierarchic_inside():
acs = ACS()
variables = ["NAME", "B01001_002E", "B01001_026E"]
geography = 'tract'
inside = 'state:01 county:001'
result = acs.data(variables, geography, inside)
result_cols = set(result.columns)
expected_cols = set(variables) | set(['tract', 'state', 'county'])
assert result_cols == expected_cols
def test_simple_data():
acs = ACS()
variables = ["NAME", "B01001_002E", "B01001_026E"]
geography = 'county'
inside = 'state:01'
result = acs.data(variables, geography, inside)
result_cols = set(result.columns)
expected_cols = {'B01001_002E', 'B01001_026E', 'NAME', 'county', 'state'}
assert result.shape[1] == 5
assert result_cols == expected_cols
def test_simple_data_dictvariables():
acs = ACS()
variables = {
'NAME': 'Geography Name',
'B01001_002E': 'Male Population',
'B01001_026E': 'Female Population',
}
geography = 'county'
inside = 'state:01'
result = acs.data(variables, geography, inside)
result_cols = set(result.columns)
expected_cols = set(variables.values()) | {'state', 'county'}
assert result_cols == expected_cols
def test_dtypes():
acs = ACS()
variables = ["NAME", "B01001_002E", "B01001_026E"]
geography = 'county'
inside = 'state:01'
df = acs.data(variables, geography, inside)
# numeric columns
result = set(c for c in df.select_dtypes('number').columns)
expected = {'B01001_002E', 'B01001_026E'}
assert result == expected
# object columns (string)
result = set(c for c in df.select_dtypes('object').columns)
expected = {'NAME', 'county', 'state'}
assert result == expected
def test_null_variables():
# When we make a request with a variable not available at the
# desired level, the Census returns nulls, and the ACS class should
# generate a warning.
acs = ACS()
variables = {
'B01001_001E': 'VARIABLE_OKAY',
'B06009_001E': 'VARIABLE_NULLS',
}
geography = 'block group'
inside = 'state:01 county:001'
with pytest.warns(UserWarning) as record:
result = acs.data(variables, geography, inside)
wmsg = record[0].message.args[0]
assert 'B06009_001E' in wmsg
assert 'B01001_001E' not in wmsg
assert set(result.columns) >= set(variables.values())
# exactly one column with all nulls
assert result.isnull().all(axis=0).sum() == 1