def test_constraint_check(self):
location = sp.load_location_from_filepath(
"test_location_grand_child.json", check_constraints=True)
self.assertEqual(location['school_size_distribution'], [0.5, 0.5])
with self.assertWarns(Warning) as wn:
sp.load_location_from_filepath("test_location_bad.json",
check_constraints=True)
self.assertTrue(
'has some negative values' in str(wn),
'Check failed: expect to get negative distribution check messages'
)
def test_check_probability_distribution_nonnegative(self, location_name='usa-Washington-seattle_metro', property_list=None):
"""
Run all checks for fields in property_list representing probability distributions. Each
should have all non negative values.
Args:
location_name(str) : name of the location json to test
property_list (list) : list of properties to check the sum of the probabilityd distribution
"""
location_file_path = f"{location_name}.json"
location = sp.load_location_from_filepath(location_file_path)
if property_list is None:
sp.logger.info(f"\nTesting all probability distributions are all non negative for {location_name}.")
checks, msgs = sp.check_all_probability_distribution_nonnegative(location)
err_msgs = [msg for msg in msgs if msg is not None]
err_msg = "\n".join(err_msgs)
assert sum(checks) == len(checks), err_msg # assert that all checks passed
print(f'All {sum(checks)} checks passed.')
else:
# Examples of how the non negative checks can be run for a subset of properties
sp.logger.info(f"\nTesting a subset of probability distributions are all non negative for {location_name}")
for i, property_name in enumerate(property_list):
check, msg = sp.check_probability_distribution_nonnegative(location, property_name)
assert check == True, msg
print(f'{property_name} check passed.')
def test_check_probability_distribution_sums(self, location_name='usa-Washington-seattle_metro', property_list=None, tolerance=1e-2):
"""
Run all checks for fields in property_list representing probability distributions. Each
should have a sum that equals 1 within the tolerance level.
Args:
location_name(str) : name of the location json to test
property_list (list) : list of properties to check the sum of the probabilityd distribution
tolerance (float) : difference from the sum of 1 tolerated
"""
location_file_path = f"{location_name}.json"
location = sp.load_location_from_filepath(location_file_path)
if property_list is None:
sp.logger.info(f"\nTesting all probability distributions sum to 1 or within tolerance {tolerance} for {location_name}.")
checks, msgs = sp.check_all_probability_distribution_sums(location, tolerance)
err_msgs = [msg for msg in msgs if msg is not None] # only get the msgs for failures
err_msg = "\n".join(err_msgs)
assert sum(checks) == len(checks), err_msg # assert that all checks passed
print(f'All {sum(checks)} checks passed.')
else:
# Example of how the sum checks can be run for a subset of properties
sp.logger.info(f"\nTesting a subset of probability distributions sum to 1 or within tolerance {tolerance} for {location_name}.")
for i, property_name in enumerate(property_list):
check, msg = sp.check_probability_distribution_sum(location, property_name, tolerance)
assert check == True, msg
print(f'{property_name} check passed.')
def setup_convert_df_to_json_array(self, pars):
"""
Set up objects to compare.
Args:
pars (dict): dictionary to get the data array and json array for comparison.
Returns:
array, json.array : An array of the desired data from a dataframe and
the json entry for comparison.
"""
df = pd.read_csv(pars.filepath)
# columns to include : include all by default
if pars.cols_ind == []:
cols = df.columns
else:
cols = df.columns[
pars.
cols_ind] # use indices to indicate which columns to include
if pars.int_cols_ind == []:
int_cols = pars.int_cols_ind
else:
int_cols = list(df.columns[pars.int_cols_ind].values)
# array-ify all the data, convert some columns to integers
arr = sp.convert_df_to_json_array(df, cols, int_cols)
# corresponding json data object for the same location and data
location = sp.load_location_from_filepath(f"{pars.location_name}.json")
json_array = getattr(location, pars.property_name)
if pars.property_name == 'population_age_distributions':
json_array = [j for j in json_array
if j.num_bins == len(arr)][0].distribution
return arr, json_array
# get the head of household age brackets and populate the household head age brackets field
get_household_head_age_brackets(location_data)
# get the head of household age distribution by household or family size and populate the household head age distribution by family size field
get_household_head_age_distribution_by_family_size(location_data)
# get the workplace size distribution array and populate the workplace size distribution field
get_workplace_size_counts_by_num_personnel(location_data)
# get age ranges by school type and populate the school types by age --- defines an age range for each school type
get_school_types_by_age(location_data)
# get reference links and populate the reference links
# you might do this step repeatedly as you find more data and populate your json
get_reference_links(location_data)
# TODO: adding notes to your json --- sometimes you may want to add additional information on how data were inferred
# you can also implement a method get_note with decorator '@location_data_construct' to return list of notes
note = ""
# add note
location_data.notes.append(note)
# save the loaded json file
sp.save_location_to_filepath(location_data, json_filepath)
print('saved.')
# check that you can reload the newly created json
new_location_data = sp.load_location_from_filepath(json_filepath)
print(f'{new_location_data.location_name} loaded.')
def test_load_minimal_location_with_parent_filepath_from_filepath(self):
child_filepath = os.path.join("unittests", "test_location_child.json")
location = sp.load_location_from_filepath(child_filepath)
self.check_minimal_location_with_parent(location)
cols=df.columns,
int_cols=['workplace_size_min', 'workplace_size_max'])
location_data.workplace_size_counts_by_num_personnel = workplace_size_dist_arr
return location_data
if __name__ == '__main__':
location_name = "Nepal"
# path to the new json we want to create for Zimbabwe
json_filepath = os.path.join(sp.settings.datadir, f'{location_name}.json')
# check if the file already exists and if not, create one
try:
location_data = sp.load_location_from_filepath(json_filepath)
except:
location_data = sp.Location()
# add the country as the location_name
location_data.location_name = location_name
# add age distribution data from raw data files
location_data = process_age_dists(location_data)
# add employment rates by age data from raw data files
location_data = process_employment_rates(location_data)
# add enrollment rates by age data from raw data files
location_data = process_enrollment_rates(location_data)
def test_load_minimal_location_with_parent_filepath_from_filepath(self):
child_filepath = "test_location_child.json"
location = sp.load_location_from_filepath(child_filepath,
check_constraints=False)
self.check_minimal_location_with_parent(location)
# We'll load location data from here.
input_location_filepath = "usa.json"
# After we modify some of the location data, we'll save it here.
output_location_filepath = "example_location.json"
# print(f'Loading location from [{input_location_filepath}], relative to synthpops config datadir: [{synthpops.config.datadir}]')
print(
f'Loading location from [{input_location_filepath}], relative to synthpops config datadir: [{sp.datadir}]'
)
# Load the location data file. When we invoke load_location_from_filepath() below, the argument will be
# interpreted relative to the directory provided to synthpops.config.set_datadir(). In this case,
# we are setting that to be the python working directory. So, the argument to load_location_from_filepath()
# will be interpreted relative to the python working directory.
location_data: sp.Location = sp.load_location_from_filepath(
input_location_filepath)
print('Modifying the location data...')
# Add a note to the notes field.
location_data.notes.append("Here's a new note added by the example code.")
# Overwrite a field. Here we have a community where kids are expected to do a lot of chores.
location_data.employment_rates_by_age = [[1, 0.00], [2, 0.00], [3, 0.00],
[4, 0.00], [5, 0.15], [6, 0.50]]
# Clear a field, irrespective of whatever it was set to before.
location_data.household_size_distribution = []
print('... done.')
# Save the location data.
