def normalize_age_and_years(
data: pd.DataFrame,
fill_value: Real = None,
cols_to_fill: List[str] = vi_globals.DRAW_COLUMNS,
gbd_round_id: int = GBD_2020_ROUND_ID,
age_group_ids: List[int] = AGE_GROUP.GBD_2020) -> pd.DataFrame:
data = vi_utils.normalize_sex(data, fill_value, cols_to_fill)
# vi_inputs.normalize_year(data)
binned_years = get_gbd_estimation_years(gbd_round_id)
years = {
'annual': list(range(min(binned_years),
max(binned_years) + 1)),
'binned': binned_years
}
if 'year_id' not in data:
# Data doesn't vary by year, so copy for each year.
df = []
for year in years['annual']:
fill_data = data.copy()
fill_data['year_id'] = year
df.append(fill_data)
data = pd.concat(df, ignore_index=True)
elif set(data.year_id) == set(years['binned']):
data = vi_utils.interpolate_year(data)
else: # set(data.year_id.unique()) == years['annual']
pass
# Dump extra data.
data = data[data.year_id.isin(years['annual'])]
data = _normalize_age(data, fill_value, cols_to_fill, age_group_ids)
return data
def test_normalize_sex_copy_3():
values = [1, 2, 3, 4]
df = pd.DataFrame({"sex_id": [3] * len(values), "value": values})
normalized = utilities.normalize_sex(df,
fill_value=0.0,
cols_to_fill=["value"])
assert (normalized.loc[normalized.sex_id == 1, "value"] == values).all()
assert (normalized.loc[normalized.sex_id == 2, "value"] == values).all()
def test_normalize_sex_fill_value():
values = [1, 2, 3, 4]
fill = 0.0
for sex in [1, 2]:
missing_sex = 1 if sex == 2 else 2
df = pd.DataFrame({"sex_id": [sex] * len(values), "value": values})
normalized = utilities.normalize_sex(df,
fill_value=fill,
cols_to_fill=["value"])
assert (normalized.loc[normalized.sex_id == sex,
"value"] == values).all()
assert (normalized.loc[normalized.sex_id == missing_sex,
"value"] == [fill] * len(values)).all()
def load_location_specific_life_expectancy(key: EntityKey, location: str):
location_id = extract.get_location_id(location)
data = extract.get_location_specific_life_expectancy(location_id)
data = data.rename(columns={'age': 'age_start'})
data['age_end'] = data.age_start.shift(-1).fillna(5.01)
earliest_year = data[data.year_id == 2025]
out = []
for year in range(project_globals.MIN_YEAR, 2025):
df = earliest_year.copy()
df['year_id'] = year
out.append(df)
data = pd.concat(out + [data], ignore_index=True)
data = utilities.normalize_sex(data, None, ['value'])
data = standardize.normalize_year(data)
data = utilities.reshape(data, value_cols=['value'])
data = utilities.scrub_gbd_conventions(data, location)
data = utilities.split_interval(data,
interval_column='year',
split_column_prefix='year')
return utilities.sort_hierarchical_data(data)
def test_normalize_sex_no_sex_id():
df = pd.DataFrame({"ColumnA": [1, 2, 3], "ColumnB": [1, 2, 3]})
normalized = utilities.normalize_sex(df,
fill_value=0.0,
cols_to_fill=["value"])
pd.testing.assert_frame_equal(df, normalized)
def test_normalize_sex(sex_ids):
df = pd.DataFrame({"sex_id": sex_ids, "value": [1] * len(sex_ids)})
normalized = utilities.normalize_sex(df,
fill_value=0.0,
cols_to_fill=["value"])
assert {1, 2} == set(normalized.sex_id)
def normalize(data: pd.DataFrame, fill_value: Real = None,
cols_to_fill: List[str] = vi_globals.DRAW_COLUMNS) -> pd.DataFrame:
data = utilities.normalize_sex(data, fill_value, cols_to_fill)
data = normalize_year(data)
data = normalize_age(data, fill_value, cols_to_fill)
return data