def test_error_report(self, mock_comparison_datframe, mock_from_data_dir):
accuracy = Accuracy(Mock(), Mock(), Mock(), Mock(), Mock(), Mock(),
Mock())
accuracy.comparison_dataframe = self._mock_comparison_dataframe()
accuracy.from_data_dir.return_value = accuracy
state_puma = dict()
state_puma['20'] = ['00500', '00602', '00604']
state_puma['29'] = ['00901', '00902']
expected_columns = ['marginal-pums', 'marginal-doppelganger']
df_puma, df_variable, df_total =\
accuracy.error_report(
state_puma, 'fake_dir',
marginal_variables=['num_people', 'num_vehicles', 'age'],
statistic=ErrorStat.ABSOLUTE_PCT_ERROR
)
# Test df_total
df_total_expected = pd.Series([2.00000, 0.666667],
index=expected_columns)
self.assertTrue(all((df_total - df_total_expected) < 1))
# Test df_puma
expected_puma_data = np.reshape([2.0, 2 / 3.0] * 5, (5, 2))
df_expected_puma = pd.DataFrame(data=expected_puma_data,
index=self._mock_state_puma(),
columns=expected_columns)
self.assertTrue((df_expected_puma == df_puma).all().all())
# Test df_variable
expected_variable_data = np.reshape([2.0, 2 / 3.0] * 12, (12, 2))
df_expected_variable = pd.DataFrame(data=expected_variable_data,
index=self._mock_variable_bins(),
columns=expected_columns)
self.assertTrue((df_expected_variable == df_variable).all().all())
# Test unimplemented statistic name
try:
self.assertRaises(
Exception,
Accuracy.error_report(
state_puma,
'fake_dir',
marginal_variables=['num_people', 'num_vehicles', 'age'],
statistic='wrong-statistic-name'))
except Exception:
pass
def population_generator():
# households_data_dirty, person_data_dirty = load_household_and_population_dfs()
pumas_to_go = set(puma_df_clean.values.tolist())
for puma in puma_df_clean:
gen_puma = 'state_{}_puma_{}_households.csv'.format(STATE, puma)
if gen_puma in os.listdir(output_dir):
print(puma)
pumas_to_go.remove(puma)
puma_tract_mappings = 'input/2010_puma_tract_mapping.txt'
configuration = Configuration.from_file('input/config.json')
preprocessor = Preprocessor.from_config(configuration.preprocessing_config)
for puma_id in pumas_to_go:
households_data = PumsData.from_csv('input/{}_household_pums_data.csv'.format(AOI_NAME)).clean(household_fields,
preprocessor,
state=str(
int(STATE)),
puma=str(int(
puma_id)))
persons_data = PumsData.from_csv('input/{}_person_pums_data.csv'.format(AOI_NAME)).clean(persons_fields,
preprocessor,
state=str(int(STATE)),
puma=str(int(puma_id)))
person_segmenter = lambda x: None
household_segmenter = lambda x: None
print("loaded")
household_model, person_model = create_bayes_net(
STATE, puma_id, output_dir,
households_data, persons_data, configuration,
person_segmenter, household_segmenter
)
marginals, allocator = download_tract_data(
STATE, puma_id, output_dir, census_api_key, puma_tract_mappings,
households_data, persons_data
)
print('Allocated {}'.format(puma_id))
population = generate_synthetic_people_and_households(
STATE, puma_id, output_dir, allocator,
person_model, household_model
)
print('Generated {}'.format(puma_id))
accuracy = Accuracy.from_doppelganger(
cleaned_data_persons=persons_data,
cleaned_data_households=households_data,
marginal_data=marginals,
population=population
)
logging.info('Absolute Percent Error for state {}, and puma {}: {}'.format(STATE, puma_id,
accuracy.absolute_pct_error().mean()))
_combine_and_synthesize()
def test_error_metrics(self, mock_comparison_dataframe):
accuracy = Accuracy(Mock(), Mock(), Mock(), Mock(), Mock(), Mock(),
Mock())
accuracy.comparison_dataframe = self._mock_comparison_dataframe()
self.assertEqual(accuracy.root_mean_squared_error(), (1.0, 1.0))
self.assertListEqual(accuracy.root_squared_error().mean().tolist(),
[1.0, 1.0])
self.assertListEqual(accuracy.absolute_pct_error().mean().tolist(),
[2.0, 0.66666666666666663])
def _generate_for_puma_data(households_raw_data, persons_raw_data,
preprocessor, puma_tract_mappings, puma_id):
households_data = households_raw_data.clean(household_fields,
preprocessor,
state=str(int(STATE)),
puma=str(int(puma_id)))
persons_data = persons_raw_data.clean(persons_fields,
preprocessor,
state=str(int(STATE)),
puma=str(int(puma_id)))
person_segmenter = lambda x: None
household_segmenter = lambda x: None
print("{} input data loaded. Starting allocation/generation.".format(
puma_id))
household_model, person_model = create_bayes_net(
STATE, puma_id, OUTPUT_DIR, households_data, persons_data,
configuration, person_segmenter, household_segmenter)
marginals, allocator = download_tract_data(STATE, puma_id, OUTPUT_DIR,
census_api_key,
puma_tract_mappings,
households_data, persons_data)
print('Allocated {}'.format(puma_id))
population = generate_synthetic_people_and_households(
STATE, puma_id, OUTPUT_DIR, allocator, person_model, household_model)
print('Generated synthetic people and households for {}'.format(puma_id))
accuracy = Accuracy.from_doppelganger(
cleaned_data_persons=persons_data,
cleaned_data_households=households_data,
marginal_data=marginals,
population=population)
print('Absolute Percent Error for state {}, and puma {}: {}'.format(
STATE, puma_id,
accuracy.absolute_pct_error().mean()))
return True
def main():
args = parse_args()
puma_tract_mappings = args.puma_tract_mappings_csv
state_id = args.state_id
puma_id = args.puma_id
census_api_key = args.census_api_key
config_file = args.config_file
output_dir = args.output_dir
db_host = args.db_host
db_database = args.db_database
db_schema = args.db_schema
db_user = args.db_user
db_password = args.db_password
configuration = Configuration.from_file(config_file)
households_data, persons_data = download_and_load_pums_data(
output_dir, state_id, puma_id, configuration, db_host, db_database,
db_schema, db_user, db_password)
household_model, person_model = create_bayes_net(
state_id, puma_id, output_dir, households_data, persons_data,
configuration, person_segmenter, household_segmenter)
marginals, allocator = download_tract_data(state_id, puma_id, output_dir,
census_api_key,
puma_tract_mappings,
households_data, persons_data)
population = generate_synthetic_people_and_households(
state_id, puma_id, output_dir, allocator, person_model,
household_model)
accuracy = Accuracy.from_doppelganger(
cleaned_data_persons=persons_data,
cleaned_data_households=households_data,
marginal_data=marginals,
population=population)
logging.info('Absolute Percent Error for state {}, and puma {}: {}'.format(
state_id, puma_id,
accuracy.absolute_pct_error().mean()))