def get_births(data, with_ntds=False):
key = 'born_with_ntds' if with_ntds else 'births'
data = pivot_data(data[project_globals.RESULT_COLUMNS(key) +
GROUPBY_COLUMNS])
data['measure'] = 'live_births_with_ntds' if with_ntds else 'live_births'
data['year'], process = data.process.str.split('_in_').str[1].str.split(
'_among_').str
data['sex'], process = process.str.split('_folic_acid_').str
# ignore the vitamin A portion, it is not relevant to birth data
data['folic_acid_fortification_group'], _ = process.str.split(
'_vitamin_a_').str
return sort_data(data.drop(columns='process'))
def aggregate_over_seed(data):
non_count_columns = []
for non_count_template in project_globals.NON_COUNT_TEMPLATES:
non_count_columns += project_globals.RESULT_COLUMNS(non_count_template)
count_columns = [
c for c in data.columns if c not in non_count_columns + GROUPBY_COLUMNS
]
non_count_data = data[non_count_columns +
GROUPBY_COLUMNS].groupby(GROUPBY_COLUMNS).mean()
count_data = data[count_columns +
GROUPBY_COLUMNS].groupby(GROUPBY_COLUMNS).sum()
return pd.concat([count_data, non_count_data], axis=1).reset_index()
def get_measure_anemia_split(data, measure):
data = pivot_data(data[project_globals.RESULT_COLUMNS(measure) +
GROUPBY_COLUMNS])
data = split_anemia_processing_column(data)
return sort_data(data)
def get_measure_birthweight_split(data, measure):
data = pivot_data(data[project_globals.RESULT_COLUMNS(measure) +
GROUPBY_COLUMNS])
data = split_birthweight_processing_column(data)
return sort_data(data.rename(columns={'process': 'measure'}))
def get_measure_no_split(data, measure):
data = pivot_data(data[project_globals.RESULT_COLUMNS(measure) +
GROUPBY_COLUMNS])
return sort_data(data.rename(columns={'process': 'measure'}))
def get_measure_data(data, measure, with_cause=True):
data = pivot_data(data[project_globals.RESULT_COLUMNS(measure) +
GROUPBY_COLUMNS])
data = split_processing_column(data, with_cause)
return sort_data(data)
def get_population_data(data):
total_pop = pivot_data(
data[[project_globals.TOTAL_POPULATION_COLUMN] +
project_globals.RESULT_COLUMNS('population') + GROUPBY_COLUMNS])
total_pop = total_pop.rename(columns={'process': 'measure'})
return sort_data(total_pop)