def gender_age_cases():
"""Fetch total country covid-19 death data."""
def _region_to_country(x):
return x\
.groupby(['date','age','sex'])\
.size()\
.reset_index(name = 'deaths')
# Poland
pl = covid19poland.covid_death_cases(from_github=True)
pl = _region_to_country(pl)
pl['country'] = 'PL'
# Czechia
cz = _src.czechia()
cz = _region_to_country(cz)
cz['country'] = 'CZ'
# match same dates
dt_range = pd.date_range(start=min(cz.date.min(), pl.date.min()),
end=min(cz.date.max(), pl.date.max()))
cz = cz[cz.date.isin(dt_range)]
pl = pl[pl.date.isin(dt_range)]
# merge
df = pd.concat([pl, cz])
return df
def country_deaths():
"""Fetch total country covid-19 death data."""
def _region_to_country(x):
return x\
.groupby(['date'])\
.size()\
.reset_index(name = 'deaths')
# Poland
pl = covid19poland.covid_death_cases(from_github=True)
pl = _region_to_country(pl)
pl['country'] = 'PL'
# Czechia
cz = _src.czechia()
cz = _region_to_country(cz)
cz['country'] = 'CZ'
# Sweden
se = sweden_resample()
se = se[['date','deaths']]\
.groupby(['date'])\
.sum()\
.reset_index()
se['country'] = 'SE'
# match same dates
dt_range = pd.date_range(start=min(cz.date.min(), pl.date.min(),
se.date.min()),
end=min(cz.date.max(), pl.date.max(),
se.date.max()))
cz = cz[cz.date.isin(dt_range)]
pl = pl[pl.date.isin(dt_range)]
se = se[se.date.isin(dt_range)]
# merge
df = pd.concat([pl, cz, se])
# fill missing
missing_rows = {'date': [], 'country': [], 'deaths': []}
for dt in dt_range:
for c in df.country.unique():
row = df[(df.date == dt) & (df.country == c)]
if row.empty:
missing_rows['date'].append(dt)
missing_rows['country'].append(c)
missing_rows['deaths'].append(0)
df = pd.concat([df, pd.DataFrame(missing_rows)], ignore_index=True)
# return
df = df\
.sort_values(['date','country'])\
.reset_index(drop = True)
return df
def test_twitter_deaths(self):
# get data
cases = PL.covid_death_cases()
x = cases.groupby(["date"]).size().reset_index(name='case_agg')
# reference
ref,_ = covid19dh.covid19("Poland", verbose = False)
ref = ref[["date","deaths"]]
ref['deaths'] = ref.deaths.diff()
# merge
x = x.merge(ref, on = "date")
# parse not matching
x = x[x.case_agg != x.deaths]
def get_covid_death_cases(self, *args, **kw):
x = PL.covid_death_cases(*args, **kw)
self.assertIsInstance(x, pd.DataFrame)
self.assertIn("date", x.columns)
self.assertIn("age", x.columns)
self.assertIn("sex", x.columns)
self.assertIn("place", x.columns)
self.assertIn("NUTS2", x.columns)
self.assertIn("NUTS3", x.columns)
self.assertIn("comorbid", x.columns)
self.assertIn("serious", x.columns)
self.assertIn("comorbid", x.columns)
return x
def get_data():
"""Get total Covid-19 data for CZ+IT+SE+PL."""
def _region_to_country(x):
"""Aggregate region data to country data."""
# aggregate
return x\
.groupby(['date','age','sex'])\
.size()\
.reset_index(name = 'deaths')
# Poland
pl = covid19poland.covid_death_cases(from_github=True)
pl = _region_to_country(pl)
pl['country'] = 'PL'
# Czechia
cz = covid19czechia.covid_deaths(level=3, usecache=True)
cz = _region_to_country(cz)
cz['country'] = 'CZ'
# Italy
it = covid19italy.covid_deaths()
it['country'] = 'IT'
# match same dates
dt_range = pd.date_range(start=min(cz.date.min(), pl.date.min()),
end=min(cz.date.max(), pl.date.max()))
cz = cz[cz.date.isin(dt_range)]
pl = pl[pl.date.isin(dt_range)]
# merge
df = pd.concat([pl, cz])
# reindex to people
df = pd.DataFrame(df.values.repeat(df.deaths.apply(int), axis=0),
columns=df.columns)
# merge
df = pd.concat([df, it])
# drop deaths column
df = df.drop(['deaths'], axis=1)
df['age'] = df.age.apply(float)
# return
return df