def plot_confirmed(region,
dates=(datetime(2020, 3, 1), datetime(2021, 3, 15)),
event=None,
save=False,
name='img/discussion/restrictions.png'):
"""Create plot of confirmed with restrictions.
Args:
region (str): Region to make the plot of.
dates (tuple): Date range (date start, date end).
event (datetime.datetime): Single event to show.
save (bool, optional): Whether to save the figure, defaultly not.
name (str, optional): Path to save the plot to.
"""
# load calendar
cal = load_calendar()
cal = cal[cal.Country == region]
if event is not None:
cal = cal[cal.Date == event]
cal['Restriction'] = cal['Title']
else:
cal = cal[(cal.Date >= dates[0]) & (cal.Date <= dates[1])]
cal['Y'] = -.05
# load data
x = posterior._posterior_data(region, dates, weekly=False)
POP = demographic.population.get_population(region)
x['Confirmed cases per 1000 people'] = x.confirmed / POP * 1e3
x = x[(x.date >= dates[0]) & (x.date <= dates[1])]
# plot
plt.rcParams["figure.figsize"] = (12, 6)
fig, ax = plt.subplots()
sns.lineplot(x='date', y='Confirmed cases per 1000 people', data=x, ax=ax)
ax.axhline(0, alpha=.4, color='grey')
sns.scatterplot(x='Date', y='Y', hue='Restriction', data=cal, ax=ax)
if save: fig.savefig(name)
def plot_weekly(dates, region, now=None, weekly=True, save=False):
"""Generates the plots of the simulation.
Args:
dates (tuple (2) of datetime): Time range for simulation (start, end).
region (str): Region to load result for.
now (datetime, optional): Time of production of the simulation. By default today.
weekly (bool, optional): Weekly data time slots if True, otherwise daily.
save (bool, optional): Whether to save the figure, defaultly not.
"""
# path
path = _get_path(dates, region, now=now, create_if_not_exists=False)
# load
x = posterior._posterior_data(region, dates, weekly=weekly)
print(x)
(sim_mean, sim_obs_mean), dt, region, params = load(dates, region, now)
print(x.shape)
print(sim_mean.shape)
# plot
posterior._plot_confirmed((sim_mean, sim_obs_mean), None, x)
if save: _savefig(f'{path}/confirmed.png')
posterior._plot_recovered((sim_mean, sim_obs_mean), None, x)
if save: _savefig(f'{path}/recovered.png')
posterior._plot_deaths((sim_mean, sim_obs_mean), None, x)
if save: _savefig(f'{path}/deaths.png')
plot_characteristics(dates, region, now, par='R0')
if save: _savefig(f'{path}/R0.png')
plot_characteristics(dates, region, now, par='Incubation period')
if save: _savefig(f'{path}/incubation.png')
plot_characteristics(dates, region, now, par='IFR')
if save: _savefig(f'{path}/ifr.png')
plot_characteristics(dates, region, now, par='Symptom duration')
if save: _savefig(f'{path}/symptom.png')
def covid_deaths(save=False, name='img/data/deaths_per100K.png'):
"""Constructs the trace plot of Covid-19 deaths.
Args:
save (bool, optional): Whether to save the figure, defaultly not.
name (str, optional): Path to save the plot to.
"""
# get data
countries = ['CZ', 'PL', 'IT', 'SE']
xx = pd.concat([
posterior._posterior_data(country,
(datetime(2020, 3, 1), datetime(2021, 5, 1)))
for country in countries
])
# population
POP = {
country: population.get_population(country)
for country in countries
}
xx['POP'] = xx.region.apply(POP.get)
# normalize
xx['deaths100K'] = xx.deaths / xx.POP * 1e5
# to weekly
xx['year'] = xx.date.apply(lambda d: int(datetime.strftime(d, '%Y')))
xx['week'] = xx.date.apply(lambda d: int(datetime.strftime(d, '%W')))
def q025(x):
return x.quantile(0.)
def q975(x):
return x.quantile(1.)
xx = xx\
.groupby(['year','week','region'])\
.aggregate({'deaths100K': 'sum'})\
.reset_index(drop=False)
xx['date'] = xx.apply(
lambda r: datetime.strptime('%04d-%02d-1' %
(r.year, r.week), '%Y-%W-%w'),
axis=1)
# plot
fig, ax = plt.subplots(figsize=(8, 6))
for label, df in xx.groupby('region'):
ax.plot(df.date, df.deaths100K, label=label)
ax.set_xlabel('Date')
ax.set_ylabel('Deaths per 100K')
ax.legend()
if save: fig.savefig(name)
def parameters(regions, delta=None, weekly=False):
"""
Args:
regions ():
delta ():
weekly ():
"""
# initialize
pars = {
'Region': [],
'Year': np.array([]),
'Month': np.array([]),
'a': np.array([]),
'c': np.array([]),
'b': np.array([]),
'd': np.array([])
}
# regions
for region in regions:
# load and crop
_, dt, params = _load_result(region=region)
if delta is None:
dates = [dt.min(), dt.max()]
else:
dates = [dt.min(), dt.min() + delta]
params = params[:, (dates[0] <= dt) & (dates[1] >= dt)]
dt = dt[(dates[0] <= dt) & (dates[1] >= dt)]
x = posterior._posterior_data(
region, (max(dt.min(), dates[0]), min(dt.max(), dates[1])),
weekly=weekly)
# dates
months = dt.apply(lambda d: int(datetime.strftime(d, '%m')))
years = dt.apply(lambda d: int(datetime.strftime(d, '%Y')))
for _ in range(dt.shape[0]):
pars['Region'].append(region)
pars['Year'] = np.concatenate([pars['Year'], years.to_numpy()])
pars['Month'] = np.concatenate([pars['Month'], months.to_numpy()])
pars['a'] = np.concatenate([pars['a'], params[0, :]])
pars['c'] = np.concatenate([pars['c'], params[1, :]])
pars['b'] = np.concatenate([pars['b'], params[2, :]])
pars['d'] = np.concatenate([pars['d'], params[3, :]])
pars = pd.DataFrame(pars)
pars['Date'] = pars.apply(lambda r: '%04d-%02d' % (r.Year, r.Month),
axis=1)
return pars
def plotSusceptible_PL_Weekly(save=False, name='TODO'):
"""Generates plot of weekly susceptible for Poland.
Args:
save (bool, optional): Whether to save the figure, defaultly not.
name (str, optional): Path to save the plot to.
"""
# load result
(sim_mean, sim_obs_mean), dates, region, params = load(
(datetime(2020, 3, 3), datetime(2021, 4, 16)), 'PL',
datetime(2021, 4, 18))
x = posterior._posterior_data(
'PL', (datetime(2020, 3, 3), datetime(2021, 4, 16)), weekly=False)
# plot
fig, ax = plt.subplots()
ax.plot(dates, sim_mean[0, :], label='S')
ax.legend()
if save: fig.savefig(name)
def _load_data(dates):
"""Load Covid-19 statistics in the appropriate format.
Args:
dates (tuple (2)): Date range of the data.
"""
global _cache
if _cache is not None:
return _cache
# regions
with open('model/regions.json') as fp:
regions = [k for k in json.load(fp) if len(k) > 2]
# fetch data
data_c, data_d, data_r = None, None, None
dateaxis = None
regions_r = []
for reg in regions:
x = posterior._posterior_data(reg, dates, weekly=True)
if dateaxis is None:
dateaxis = x.date
POP = population.get_population(reg)
# normalize by popylation
x['I1K'] = x.confirmed / POP * 1e3
x['D1K'] = x.deaths / POP * 1e3
# confirmed
c = x.I1K.to_numpy().reshape((1, -1))
data_c = np.concatenate([data_c, c],
axis=0) if data_c is not None else c
# deaths
d = x.D1K.to_numpy().reshape((1, -1))
data_d = np.concatenate([data_d, d],
axis=0) if data_d is not None else d
# recovered
if 'recovered' in x:
x['R1K'] = x.recovered / POP * 1e3
r = x.recovered.to_numpy().reshape((1, -1))
data_r = np.concatenate([data_r, r],
axis=0) if data_r is not None else r
regions_r.append(reg)
_cache = (data_c, data_d, data_r), dateaxis, (regions, regions, regions_r)
return _cache
def covid_recovered(save=False, name='img/data/recovered_per100K.png'):
"""Constructs the trace plot of Covid-19 recovered.
Args:
save (bool, optional): Whether to save the figure, defaultly not.
name (str, optional): Path to save the plot to.
"""
# get data
countries = ['CZ', 'PL', 'IT']
xx = pd.concat([
posterior._posterior_data(country,
(datetime(2020, 3, 1), datetime(2021, 5, 1)))
for country in countries
])
# population
POP = {
country: population.get_population(country)
for country in countries
}
xx['POP'] = xx.region.apply(POP.get)
# normalize
xx['deaths100K'] = xx.deaths / xx.POP * 1e5
# to weekly
xx['year'] = xx.date.apply(lambda d: int(datetime.strftime(d, '%Y')))
xx['week'] = xx.date.apply(lambda d: int(datetime.strftime(d, '%W')))
def q025(x):
return x.quantile(0.)
def q975(x):
return x.quantile(1.)
# plot
fig, ax = plt.subplots(figsize=(8, 6))
for label, df in xx.groupby('region'):
alpha = 1 if label != 'PL' else .5
ax.plot(df.date, df.deaths100K, label=label, alpha=alpha)
ax.set_xlabel('Date')
ax.set_ylabel('Deaths per 100K')
ax.legend()
if save: fig.savefig(name)
def prediction_data_correlation(regions, components, delta=None, weekly=False):
"""
Args:
regions ():
components ():
delta ():
weekly ():
"""
# initialize
corrs = {'region': regions}
for c in components:
corrs[c] = []
# regions
for region in regions:
# load and crop
lat, dt, _ = _load_result(region=region)
if delta is None:
dates = [dt.min(), dt.max()]
else:
dates = [dt.min(), dt.min() + delta]
lat = lat[:, (dates[0] <= dt) & (dates[1] >= dt)]
dt = dt[(dates[0] <= dt) & (dates[1] >= dt)]
x = posterior._posterior_data(
region, (max(dt.min(), dates[0]), min(dt.max(), dates[1])),
weekly=weekly)
# compute correlation
if 'I' in components:
corrs['I'].append(
np.corrcoef(lat[2, :], x.confirmed.to_numpy())[1, 0])
if 'R' in components:
corrs['R'].append(
np.corrcoef(lat[3, :], x.recovered.to_numpy())[1, 0])
if 'D' in components:
corrs['D'].append(
np.corrcoef(lat[4, :], x.deaths.to_numpy())[1, 0])
corrs = pd.DataFrame(corrs)
return corrs
def save(sim, dates, region, params):
"""Save results to file.
Args:
sim (tuple (2) of matrices): Simulation results (latent [5xD], observed [5xD]).
dates (tuple (2) of datetime): Time range for simulation (start, end).
region (str): Region to load result for.
now (datetime, optional): Time of production of the simulation. By default today.
"""
# path
path = _get_path(dates, region)
# parse simulations
sim_lat, sim_obs = sim
lat = sim_lat.mean(axis=1)
obs = sim_obs.mean(axis=1)
x = posterior._posterior_data(region, dates)
# save
df = pd.DataFrame({
'date': x.date,
'region': region,
'param_a': None, 'param_c': None, 'param_b': None, 'param_d': None,
'latent_S': lat[0,:],
'latent_E': lat[1,:],
'latent_I': lat[2,:],
'latent_R': lat[3,:],
'latent_D': lat[4,:],
'observed_I': obs[2,:], 'observed_R': obs[3,:], 'observed_D': obs[4,:],
'data_I': x.confirmed,
'data_R': x.recovered if 'recovered' in x.columns else np.nan,
'data_D': x.deaths})\
.reset_index(drop=True)
# set parameters a,c,b,d
for df_i in df.index:
df_date = df.loc[df_i, 'date']
params_i = (params.start <= df_date) & (params.end >= df_date)
params_i = params[params_i].start.idxmax()
for col in ['a', 'c', 'b', 'd']:
df.loc[df_i, 'param_' + col] = float(params.loc[params_i, col])
# save
df.to_csv(f'{path}/data.csv', index=False)
# aggregate
sim_mean, sim_obs_mean = posterior.compute_sim_mean(sim)
sim_ci, sim_obs_ci = posterior.compute_sim_ci(sim)
# plot
posterior._plot_confirmed((sim_mean, sim_obs_mean), (sim_ci, sim_obs_ci),
x)
plt.savefig(f'{path}/confirmed.png')
posterior._plot_recovered((sim_mean, sim_obs_mean), (sim_ci, sim_obs_ci),
x)
plt.savefig(f'{path}/recovered.png')
posterior._plot_deaths((sim_mean, sim_obs_mean), (sim_ci, sim_obs_ci), x)
plt.savefig(f'{path}/deaths.png')
plot_characteristics(dates, region, datetime.now(), par='R0')
plt.savefig(f'{path}/R0.png')
plot_characteristics(dates,
region,
datetime.now(),
par='Incubation period')
plt.savefig(f'{path}/incubation.png')
plot_characteristics(dates, region, datetime.now(), par='IFR')
plt.savefig(f'{path}/ifr.png')
plot_characteristics(dates, region, datetime.now(), par='Symptom duration')
plt.savefig(f'{path}/symptom.png')
