DEATH_COUNT_THRESHOLD = 10
COUNTRIES_CORE = 'IT DE UK ES IE FR'.split()
COUNTRIES_NORDIC = 'SE NO DK FI UK'.split()
-COUNTRIES_FRIENDS = 'IT UK ES BE SI MX'.split()
+# COUNTRIES_FRIENDS = 'IT UK ES BE SI MX'.split()
+COUNTRIES_FRIENDS = 'IT UK ES BE SI PT'.split()
+
+COUNTRIES_AMERICAS = ['AG', 'AR', 'AW', 'BS', 'BB', 'BZ', 'BM', 'BO', 'BR', 'VG', 'KY', # excluding Canada and USA
+ 'CL', 'CO', 'CR', 'CU', 'CW', 'DM', 'DO', 'EC', 'SV', 'GL', 'GD', 'GT',
+ 'GY', 'HT', 'HN', 'JM', 'MX', 'MS', 'NI', 'PA', 'PY', 'PE', 'PR', 'KN',
+ 'LC', 'VC', 'SX', 'SR', 'TT', 'TC', 'VI', 'UY', 'VE']
COUNTRIES_OF_INTEREST = list(set(COUNTRIES_CORE + COUNTRIES_FRIENDS))
-COUNTRIES_ALL = list(set(COUNTRIES_CORE + COUNTRIES_FRIENDS + COUNTRIES_NORDIC))
+COUNTRIES_ALL = list(set(COUNTRIES_CORE + COUNTRIES_FRIENDS + COUNTRIES_NORDIC + COUNTRIES_AMERICAS))
```
```python
```python
# First col is a date, treat geoId of NA as 'Namibia', not "NA" value
-raw_data = pd.read_csv('covid.csv', parse_dates=[0], keep_default_na=False, dayfirst=True)
+raw_data = pd.read_csv('covid.csv',
+ parse_dates=[0], dayfirst=True,
+ keep_default_na=False, na_values = [''],
+# dtype = {'day': np.int64,
+# 'month': np.int64,
+# 'year': np.int64,
+# 'cases': np.int64,
+# 'deaths': np.int64,
+# 'countriesAndTerritories': str,
+# 'geoId': str,
+# 'countryterritoryCode': str,
+# 'popData2019': np.int64,
+# 'continentExp': str,
+# }
+ )
```
```python
raw_data.size
```
+```python
+raw_data.fillna(0, inplace=True)
+```
+
```python
raw_data.head()
```
```
```python
-base_data = raw_data.set_index(['geoId', 'dateRep'])
-base_data.sort_index(inplace=True)
-base_data
+raw_data = raw_data.astype({'dateRep': np.datetime64,
+ 'day': np.int64,
+ 'month': np.int64,
+ 'year': np.int64,
+ 'cases': np.int64,
+ 'deaths': np.int64,
+ 'countriesAndTerritories': str,
+ 'geoId': str,
+ 'countryterritoryCode': str,
+ 'popData2019': np.int64,
+ 'continentExp': str })
+```
+
+```python
+raw_data.dtypes
```
```python
-base_data.loc['UK']
+raw_data[((raw_data.geoId == 'UK') & (raw_data.dateRep >= '2020-07-10'))]
```
```python
-base_data.loc['UK', '2020-04-17']
+# raw_data = raw_data[~ ((raw_data.geoId == 'ES') & (raw_data.dateRep >= '2020-05-22'))]
```
```python
-countries = raw_data[['geoId', 'countriesAndTerritories', 'popData2018']]
-countries = countries[countries['popData2018'] != '']
+base_data = raw_data.set_index(['geoId', 'dateRep'])
+base_data.sort_index(inplace=True)
+base_data
+```
+
+```python
+base_data.loc['ES'].loc['2020-05-10':]
+```
+
+```python
+countries = raw_data[['geoId', 'countriesAndTerritories', 'popData2019', 'continentExp']]
+countries = countries[countries['popData2019'] != '']
countries = countries.drop_duplicates()
countries.set_index('geoId', inplace=True)
-countries = countries.astype({'popData2018': 'int64'})
+countries = countries.astype({'popData2019': 'int64'})
countries.head()
```
countries.loc[COUNTRIES_OF_INTEREST]
```
+```python
+countries[countries.continentExp == 'America'].index
+```
+
```python
data_by_date = base_data[['cases', 'deaths']]
data_by_date.head()
data_by_date.loc['UK']
```
+```python
+# data_by_date.deaths.drop_duplicates().sort_values().to_csv('dth.csv', header=True)
+```
+
```python
data_by_date.groupby(level=0).cumsum()
```
data_since_threshold = data_since_threshold.set_index('since_threshold', append=True
).reorder_levels(['since_threshold', 'geoId', 'dateRep']
).reset_index('dateRep')
+data_since_threshold.sort_index(inplace=True)
data_since_threshold
```
data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), :]
```
+```python
+data_since_threshold.loc[(slice(None), ['ES']), :].tail(8)
+```
+
```python
data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), ['deaths_culm']].unstack().plot(logy=True)
```
deaths = data_since_threshold.loc[(slice(None), COUNTRIES_ALL), ['deaths_culm']].unstack().sort_index().xs('deaths_culm', axis=1, drop_level=True)
```
+```python
+cases = data_since_threshold.loc[(slice(None), COUNTRIES_ALL), ['cases_culm']].unstack().sort_index().xs('cases_culm', axis=1, drop_level=True)
+```
+
+```python
+COUNTRIES_AMERICAS_DEAD = list(set(deaths.columns) & set(COUNTRIES_AMERICAS))
+```
+
```python
data_since_threshold.reset_index().merge(countries, on='geoId').set_index(['since_threshold', 'geoId'])
```
```python
data_since_threshold_per_capita = data_since_threshold.reset_index().merge(countries, on='geoId').set_index(['since_threshold', 'geoId'])
-data_since_threshold_per_capita['cases_culm_pc'] = data_since_threshold_per_capita.cases_culm / data_since_threshold_per_capita.popData2018
-data_since_threshold_per_capita['deaths_culm_pc'] = data_since_threshold_per_capita.deaths_culm / data_since_threshold_per_capita.popData2018
+data_since_threshold_per_capita['cases_culm_pc'] = data_since_threshold_per_capita.cases_culm / data_since_threshold_per_capita.popData2019
+data_since_threshold_per_capita['deaths_culm_pc'] = data_since_threshold_per_capita.deaths_culm / data_since_threshold_per_capita.popData2019
data_since_threshold_per_capita
```
deaths_pc = data_since_threshold_per_capita.loc[(slice(None), ['UK', 'DE', 'IT', 'IE']), ['deaths_culm_pc']].unstack().sort_index().xs('deaths_culm_pc', axis=1, drop_level=True)
```
-```python
-deaths_pc
-```
-
```python
deaths_pc.index
```
deaths[COUNTRIES_FRIENDS].plot()
```
+```python
+ax = deaths[COUNTRIES_FRIENDS].plot(figsize=(10, 6), title="Total deaths, linear")
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+for c in COUNTRIES_FRIENDS:
+ lvi = deaths[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths[c][lvi], s = f"{c}: {deaths[c][lvi]:.0f}")
+# plt.savefig('covid_deaths_total_linear.png')
+```
+
```python
ax = deaths[COUNTRIES_CORE].plot(figsize=(10, 6), title="Total deaths, linear")
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
for c in COUNTRIES_CORE:
lvi = deaths[c].last_valid_index()
- ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
+ ax.text(x = lvi + 1, y = deaths[c][lvi], s = f"{c}: {deaths[c][lvi]:.0f}")
plt.savefig('covid_deaths_total_linear.png')
```
+```python
+deaths_prime = deaths[COUNTRIES_CORE].copy()
+deaths_prime.loc[73:, 'ES'] = np.NaN
+# deaths_prime['ES'][70:]
+```
+
+```python
+ax = deaths_prime[COUNTRIES_CORE].plot(figsize=(10, 6), title="Total deaths, linear")
+for c in COUNTRIES_CORE:
+ lvi = deaths_prime[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths_prime[c][lvi], s = f"{c}: {deaths_prime[c][lvi]:.0f}")
+# plt.savefig('covid_deaths_total_linear.png')
+```
+
+```python
+ax = cases[COUNTRIES_CORE].plot(figsize=(10, 6), title="Total cases, linear")
+for c in COUNTRIES_CORE:
+ lvi = cases[c].last_valid_index()
+ ax.text(x = lvi + 1, y = cases[c][lvi], s = c)
+plt.savefig('covid_cases_total_linear.png')
+```
+
+```python
+ax = deaths[COUNTRIES_AMERICAS_DEAD].plot(figsize=(10, 6), title="Total deaths, linear")
+for c in COUNTRIES_AMERICAS_DEAD:
+ lvi = deaths[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
+# plt.savefig('covid_deaths_total_linear.png')
+```
+
+```python
+ax = deaths[COUNTRIES_CORE + ['BR', 'MX']].plot(figsize=(10, 6), title="Total deaths, linear")
+for c in COUNTRIES_CORE + ['BR', 'MX']:
+ lvi = deaths[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
+# plt.savefig('covid_deaths_total_linear.png')
+```
+
```python
ax = deaths[COUNTRIES_NORDIC].plot(figsize=(10, 6), title="Total deaths, linear")
for c in COUNTRIES_NORDIC:
for c in COUNTRIES_OF_INTEREST:
lvi = deaths[c].last_valid_index()
ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
-# plt.savefig('covid_deaths_total_linear.png')
+plt.savefig('covid_deaths_total_linear_of_interest.png')
```
```python
```python
data_since_threshold['deaths_m4'] = data_since_threshold.groupby(level=1)['deaths'].transform(lambda x: x.rolling(4, 1).mean())
data_since_threshold['deaths_m7'] = data_since_threshold.groupby(level=1)['deaths'].transform(lambda x: x.rolling(7, 1).mean())
+data_since_threshold['cases_m7'] = data_since_threshold.groupby(level=1)['cases'].transform(lambda x: x.rolling(7, 1).mean())
# data_since_threshold['deaths_diff_m4'] = data_since_threshold.groupby(level=1)['deaths_diff'].transform(lambda x: x.rolling(4, 1).mean())
# data_since_threshold['deaths_diff_m7'] = data_since_threshold.groupby(level=1)['deaths_diff'].transform(lambda x: x.rolling(7, 1).mean())
data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), :]
.unstack().sort_index().xs('deaths_m7', axis=1, drop_level=True))
```
+```python
+cases_m7 = (data_since_threshold.loc[(slice(None), COUNTRIES_ALL), ['cases_m7']]
+ .unstack().sort_index().xs('cases_m7', axis=1, drop_level=True))
+```
+
```python
ax = deaths_m4.plot(figsize=(10, 6), title="Deaths per day, 4 day moving average")
for c in deaths_m4.columns:
```python
ax = deaths_m7[COUNTRIES_CORE].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average")
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
for c in COUNTRIES_CORE:
lvi = deaths_m7[c].last_valid_index()
ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = c)
+# plt.axhline(0, color='0.7')
plt.savefig('covid_deaths_per_day_7.png')
```
+```python
+ax = deaths_m7[COUNTRIES_FRIENDS].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average")
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+for c in COUNTRIES_FRIENDS:
+ lvi = deaths_m7[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = c)
+# plt.axhline(0, color='0.7')
+# plt.savefig('covid_deaths_per_day_7.png')
+```
+
+```python
+deaths_m7_prime = deaths_m7[COUNTRIES_CORE].copy()
+deaths_m7_prime.loc[73:, 'ES'] = np.NaN
+deaths_m7_prime['ES'][70:]
+```
+
+```python
+ax = deaths_m7_prime[COUNTRIES_CORE].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average")
+for c in COUNTRIES_CORE:
+ lvi = deaths_m7_prime[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths_m7_prime[c][lvi], s = c)
+# plt.savefig('covid_deaths_per_day_7.png') # see below for where this is written, with the projection
+```
+
```python
ax = deaths_m7[COUNTRIES_FRIENDS].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average")
for c in COUNTRIES_FRIENDS:
plt.savefig('covid_deaths_per_day_friends_7.png')
```
+```python
+ax = deaths_m7[COUNTRIES_CORE + ['BR', 'MX']].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average")
+for c in COUNTRIES_CORE + ['BR', 'MX']:
+ lvi = deaths_m7[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = c)
+# plt.savefig('covid_deaths_per_day_7.png')
+```
+
+```python
+ax = cases_m7[COUNTRIES_CORE].plot(figsize=(10, 6), title="Cases per day, 7 day moving average")
+for c in COUNTRIES_CORE:
+ lvi = cases_m7[c].last_valid_index()
+ ax.text(x = lvi + 1, y = cases_m7[c][lvi], s = c)
+plt.savefig('covid_cases_per_day-core.png')
+```
+
+```python
+ax = cases_m7[COUNTRIES_FRIENDS].plot(figsize=(10, 6), title="Cases per day, 7 day moving average")
+for c in COUNTRIES_FRIENDS:
+ lvi = cases_m7[c].last_valid_index()
+ ax.text(x = lvi + 1, y = cases_m7[c][lvi], s = c)
+# plt.savefig('covid_cases_per_day-core.png')
+```
+
```python
def gmean_scale(items):
return gmean(items) / items[-1]
```
```python
-data_since_threshold['doubling_time'] = data_since_threshold.groupby(level=1).apply(doubling_time).reset_index(level=0, drop=True)
-data_since_threshold['doubling_time_7'] = data_since_threshold.groupby(level=1).apply(doubling_time_7).reset_index(level=0, drop=True)
+data_since_threshold['doubling_time'] = data_since_threshold.groupby(level=1).apply(doubling_time).reset_index(level=0, drop=True).sort_index()
+data_since_threshold['doubling_time_7'] = data_since_threshold.groupby(level=1).apply(doubling_time_7).reset_index(level=0, drop=True).sort_index()
# data_since_threshold.loc[(slice(None), 'UK'), :]
```
```python
ax = doubling_times_7[COUNTRIES_CORE].plot(figsize=(10, 6), title="Doubling times, 7 day average")
+ax.legend(loc="upper left")
for c in COUNTRIES_CORE:
lvi = doubling_times_7[c].last_valid_index()
ax.text(x = lvi + 1, y = doubling_times_7[c][lvi], s = c)
```
```python
-data_since_threshold.replace([np.inf, -np.inf], np.nan).groupby(level=1).last().loc[COUNTRIES_ALL]#, [doubling_time]]
+# data_since_threshold.replace([np.inf, -np.inf], np.nan).groupby(level=1).last().loc[COUNTRIES_ALL]#, [doubling_time]]
+```
+
+```python
+dstl = data_since_threshold.replace([np.inf, -np.inf], np.nan).groupby(level=1).last()
+dstl.loc[dstl.index.intersection(COUNTRIES_ALL)]
```
```python
-data_since_threshold.replace([np.inf, -np.inf], np.nan).groupby(level=1).last().loc[['UK', 'DE', 'IT']]#, [doubling_time]]
+# data_since_threshold.replace([np.inf, -np.inf], np.nan).groupby(level=1).last().loc[['UK', 'DE', 'IT']]#, [doubling_time]]
+dstl.loc[['UK', 'DE', 'IT', 'FR', 'ES']]
+```
+
+```python
+data_since_threshold.loc[(slice(None), ['UK']), :].tail(20)
+```
+
+```python
+data_since_threshold.loc[(slice(None), ['ES']), :].tail(20)
+```
+
+## Death projections
+
+```python
+data_since_threshold.loc[(slice(None), ['UK']), :].tail(15)
```
```python
```python
uk_projection = data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['UK']), :]
-s_start = uk_projection.index.max()[0] + 1
+uk_current_end = uk_projection.index.max()[0] + 1
+# s_start = uk_projection.index.max()[0] + 1
+uk_current_end
+```
+
+```python
+current_uk_deaths_m7 = uk_projection[uk_projection.deaths_m7 >= 0].iloc[-1].deaths_m7
+current_uk_deaths_m7
+```
+
+```python
+it_since_threshold[it_since_threshold.deaths_m7 <= current_uk_deaths_m7].loc[60:].first_valid_index()[0]
+```
+
+```python
+s_start = it_since_threshold[it_since_threshold.deaths_m7 <= current_uk_deaths_m7].loc[60:].first_valid_index()[0]
s_start
```
```python
-proj = it_since_threshold.loc[(slice(s_start, s_end), slice(None)), ['cases', 'deaths']]
-proj.index = pd.MultiIndex.from_tuples([(n, 'UK') for n, _ in proj.index], names=proj.index.names)
+s_start_date = data_since_threshold.loc[(89, 'IT'), 'dateRep']# .iloc[0]
+s_start_date
+```
+
+```python
+s_end - s_start
+```
+
+```python
+uk_end = s_end - s_start + uk_current_end
+uk_end
+```
+
+```python
+proj = it_since_threshold.loc[(slice(s_start, s_end), slice(None)), ['cases', 'deaths', 'deaths_m7']]
+ndiff = uk_current_end - s_start
+proj.index = pd.MultiIndex.from_tuples([(n + ndiff, 'UK') for n, _ in proj.index], names=proj.index.names)
proj
```
+```python
+it_since_threshold.loc[(slice(s_start - 8, s_start + 2), slice(None)), ['cases', 'deaths', 'deaths_m7']]
+```
+
+```python
+uk_projection[['cases', 'deaths', 'deaths_m7']].tail()
+```
+
+```python
+# proj['deaths_m7'] = proj['deaths_m7'] + 20
+# proj
+```
+
Projected deaths, UK following IT trend from now.
```python
uk_projection.deaths.sum()
```
+```python
+uk_projection = uk_projection.droplevel(1)
+uk_projection
+```
+
+```python
+uk_projection.loc[152, 'deaths']
+```
+
+## Correction for cumulative deaths correction on 14 August
+
+```python
+# uk_projection.loc[152, 'deaths'] = 50
+```
+
+```python
+uk_projection['deaths_m7'] = uk_projection['deaths'].transform(lambda x: x.rolling(7, 1).mean())
+uk_projection.loc[(uk_current_end - 20):(uk_current_end + 5)]
+```
+
+```python
+uk_projection.loc[(uk_current_end - 5):]
+```
+
+```python
+uk_projection.deaths_m7.plot()
+```
+
+```python
+proj.droplevel(level=1)
+```
+
+```python
+ax = deaths_m7[COUNTRIES_CORE].plot()
+# uk_projection['deaths_m7'].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average", label="Projection", style='--', ax=ax)
+proj.droplevel(level=1)['deaths_m7'].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average", label="Projection", style='--', ax=ax)
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+for c in COUNTRIES_CORE:
+ lvi = deaths_m7[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = c)
+plt.savefig('covid_deaths_per_day_7.png')
+```
+
```python
it_since_threshold.deaths.sum()
```
+# Excess deaths calculation
+
```python
with open('excess_deaths.json') as f:
excess_deaths_data = json.load(f)
-excess_deaths_data
+
+with open('excess_death_accuracy.json') as f:
+ excess_death_accuracy = json.load(f)
+
+excess_deaths_data, excess_death_accuracy
```
```python
-excess_deaths_upto = '2020-05-08'
-excess_deaths = 54500
+# excess_deaths_upto = '2020-05-08'
+# excess_deaths = 54500
```
```python
Recorded deaths in period where ONS has reported total deaths
```python
-reported_deaths = base_data.loc['UK'][:excess_deaths_upto]['deaths'].sum()
-reported_deaths
+ons_reported_deaths = base_data.loc['UK'][:excess_deaths_upto]['deaths'].sum()
+ons_reported_deaths
+```
+
+```python
+excess_deaths_upto
+```
+
+## Correction for deaths total correction on 14 August
+
+```python
+ons_unreported_deaths_data = base_data.loc['UK'][excess_deaths_upto:].iloc[1:]['deaths']
+# ons_unreported_deaths_data['2020-08-14'] = 50
```
```python
-excess_deaths / reported_deaths
+ons_unreported_deaths = ons_unreported_deaths_data.sum()
+ons_unreported_deaths
+```
+
+```python
+scaled_ons_unreported_deaths = ons_unreported_deaths * excess_death_accuracy
+scaled_ons_unreported_deaths
+```
+
+```python
+uk_deaths_to_date = excess_deaths + scaled_ons_unreported_deaths
+uk_deaths_to_date
+```
+
+```python
+# data_since_threshold.loc[(slice(None), 'UK'), :][data_since_threshold.dateRep == excess_deaths_data['end_date']]
+```
+
+```python
+data_since_threshold[data_since_threshold.dateRep == excess_deaths_data['end_date']].loc[(slice(None), 'UK'), :]
+```
+
+```python
+ons_unreported_start = data_since_threshold[data_since_threshold.dateRep == excess_deaths_data['end_date']].loc[(slice(None), 'UK'), :].first_valid_index()[0] + 1
+ons_unreported_start
+```
+
+```python
+unreported_projected_deaths = uk_projection.loc[ons_unreported_start:].deaths.sum()
+unreported_projected_deaths
+```
+
+```python
+unreported_projected_deaths_scaled = unreported_projected_deaths * excess_death_accuracy
+unreported_projected_deaths_scaled
+```
+
+```python
+uk_projection.loc[(s_start):].deaths.sum()
+```
+
+```python
+deaths_actual_projected_scaled = uk_deaths_to_date + uk_projection.loc[(s_start):].deaths.sum() * excess_death_accuracy
+deaths_actual_projected_scaled
+```
+
+```python
+# excess_deaths / reported_deaths
```
True deaths to date, if we follow the scaling of excess deaths over reported deaths so far.
```
```python
-data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['IT']), 'dateRep'].iloc[-1] + pd.Timedelta(s_end - s_start, unit='days')
+# uk_covid_deaths_scaled = excess_deaths + unreported_deaths * excess_death_accuracy
+# uk_covid_deaths_scaled
+```
+
+```python
+# data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['IT']), 'dateRep'].iloc[-1] + pd.Timedelta(s_end - s_start, unit='days')
+```
+
+```python
+# data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['UK']), 'dateRep'].iloc[-1].strftime("%Y-%m-%d")
+```
+
+```python
+# uk_covid_deaths * excess_deaths / reported_deaths
```
```python
-data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['UK']), 'dateRep'].iloc[-1].strftime("%Y-%m-%d")
+# uk_projection.deaths.sum() * excess_deaths / reported_deaths
```
```python
-uk_covid_deaths * excess_deaths / reported_deaths
+# data_since_threshold.loc[(slice(None), 'FR'), :]
+# data_since_threshold[data_since_threshold.dateRep == '2020-05-18'].loc[(slice(None), 'FR'), :]
```
+## School reopenings
+
```python
-uk_projection.deaths.sum() * excess_deaths / reported_deaths
+school_reopenings = {
+ 'ES': {'date': '2020-05-18'},
+ 'FR': {'date': '2020-05-18'}, # some areas only
+# 'IT': {'date': '2020-09-01'},
+ # 'IE': {'date': '2020-09-01'},
+ 'DE': {'date': '2020-05-04'},
+ 'UK': {'date': '2020-06-01'}
+}
+```
+
+```python
+data_since_threshold[data_since_threshold.dateRep == '2020-05-04'].loc[(slice(None), ['DE']), :].first_valid_index()
+```
+
+```python
+data_since_threshold[data_since_threshold.dateRep == '2020-05-04'].loc[(slice(None), ['DE']), :].iloc[0].deaths_m7
+```
+
+```python
+for cID in school_reopenings:
+ dst_in = data_since_threshold[data_since_threshold.dateRep == (school_reopenings[cID]['date'])].loc[(slice(None), [cID]), :]
+ dst_i = dst_in.first_valid_index()
+ dst_n = dst_in.iloc[0].deaths_m7
+ school_reopenings[cID]['since_threshold'] = dst_i[0]
+ school_reopenings[cID]['deaths_m7'] = dst_n
+school_reopenings
+```
+
+```python
+ax = deaths_m7[COUNTRIES_CORE].plot(figsize=(15, 9), title="Deaths per day, 7 day moving average")
+# uk_projection.deaths_m7.plot(ax=ax)
+for c in COUNTRIES_CORE:
+ lvi = deaths_m7[c].last_valid_index()
+ ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = f"{c}: {deaths_m7[c][lvi]:.0f}")
+ if c in school_reopenings:
+ marker_col = [l for l in ax.lines if l.get_label() == c][0].get_color()
+ ax.plot(school_reopenings[c]['since_threshold'], school_reopenings[c]['deaths_m7'], '*',
+ markersize=18, markerfacecolor=marker_col, markeredgecolor=marker_col)
+ ax.text(x = school_reopenings[c]['since_threshold'] + 1, y = school_reopenings[c]['deaths_m7'],
+ s = f"{school_reopenings[c]['date']}: {school_reopenings[c]['deaths_m7']:.0f}")
+plt.savefig('school_reopenings.png')
+```
+
+```python
+# ax = deaths_m7[COUNTRIES_CORE].plot(figsize=(15, 9), title="Deaths per day, 7 day moving average",
+# xlim=(46, 91), ylim=(0, 400))
+# # uk_projection.deaths_m7.plot(ax=ax)
+# for c in COUNTRIES_CORE:
+# lvi = deaths_m7[c].last_valid_index()
+# ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = f"{c}: {deaths_m7[c][lvi]:.0f}", fontsize=14)
+# if c in school_reopenings:
+# marker_col = [l for l in ax.lines if l.get_label() == c][0].get_color()
+# ax.plot(school_reopenings[c]['since_threshold'], school_reopenings[c]['deaths_m7'], '*',
+# markersize=18, markerfacecolor=marker_col, markeredgecolor=marker_col)
+# ax.text(x = school_reopenings[c]['since_threshold'] + 1, y = school_reopenings[c]['deaths_m7'],
+# s = f"{school_reopenings[c]['date']}: {school_reopenings[c]['deaths_m7']:.0f}",
+# fontsize=14)
+# plt.savefig('school_reopenings_detail.png')
+```
+
+# Lockdown graphs
+
+```python
+lockdown_dates = {
+ 'ES': { 'part_start': {'date': '2020-03-14'}
+ , 'full_start': {'date': '2020-03-15'}
+ , 'part_finish': {'date': '2020-05-18'}
+ },
+ 'FR': { 'part_start': {'date': '2020-03-13'}
+ , 'full_start': {'date': '2020-03-17'}
+ , 'part_finish': {'date': '2020-05-11'}
+ },
+ 'IT': { 'part_start': {'date': '2020-03-08'}
+ , 'full_start': {'date': '2020-03-10'}
+ , 'part_finish': {'date': '2020-05-04'}
+ },
+ 'DE': { #'part_start': {'date': '2020-03-13'}
+ 'full_start': {'date': '2020-03-22'}
+ , 'part_finish': {'date': '2020-05-06'}
+ },
+ 'UK': { 'part_start': {'date': '2020-03-23'}
+ , 'full_start': {'date': '2020-03-23'}
+ , 'part_finish': {'date': '2020-05-31'}
+ },
+ 'IE': { #'part_start': {'date': '2020-03-12'}
+ 'full_start': {'date': '2020-03-27'}
+ , 'part_finish': {'date': '2020-05-18'}
+ },
+}
+```
+
+```python
+for cID in lockdown_dates:
+ for phase in lockdown_dates[cID]:
+ dst_in = data_since_threshold[data_since_threshold.dateRep == (lockdown_dates[cID][phase]['date'])].loc[(slice(None), [cID]), :]
+ dst_i = dst_in.first_valid_index()
+ dst_n = dst_in.iloc[0].deaths_m7
+ dst_c = dst_in.iloc[0].cases_m7
+ lockdown_dates[cID][phase]['since_threshold'] = dst_i[0]
+ lockdown_dates[cID][phase]['deaths_m7'] = dst_n
+ lockdown_dates[cID][phase]['cases_m7'] = dst_c
+
+lockdown_dates
+```
+
+```python
+ax = deaths_m7[COUNTRIES_CORE].plot(figsize=(15, 9), title="Deaths per day, 7 day moving averagee, with lockdown dates")
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+# uk_projection.deaths_m7.plot(ax=ax)
+for c in COUNTRIES_CORE:
+ lvi = deaths_m7[c].last_valid_index()
+ if c != 'UK':
+ ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = f"{c}: {deaths_m7[c][lvi]:.0f}")
+ if c in lockdown_dates:
+ for phase in lockdown_dates[c]:
+ marker_col = [l for l in ax.lines if l.get_label() == c][0].get_color()
+ ax.plot(lockdown_dates[c][phase]['since_threshold'], lockdown_dates[c][phase]['deaths_m7'], '*',
+ markersize=18, markerfacecolor=marker_col, markeredgecolor=marker_col)
+ if 'start' not in phase:
+ ax.text(x = lockdown_dates[c][phase]['since_threshold'] + 1, y = lockdown_dates[c][phase]['deaths_m7'],
+ s = f"{lockdown_dates[c][phase]['date']}: {lockdown_dates[c][phase]['deaths_m7']:.0f}")
+# plt.savefig('school_reopenings.png')
+```
+
+```python
+ax = cases_m7.iloc[-50:][COUNTRIES_CORE].plot(figsize=(15, 9), title="Cases per day, 7 day moving average, with lockdown dates") #, ylim=(-10, 1500))
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+# uk_projection.deaths_m7.plot(ax=ax)
+for c in COUNTRIES_CORE:
+ lvi = cases_m7[c].last_valid_index()
+# if c != 'UK':
+ ax.text(x = lvi + 1, y = cases_m7[c][lvi], s = f"{c}: {cases_m7[c][lvi]:.0f}")
+
+```
+
+```python
+ax = cases_m7[COUNTRIES_CORE].plot(figsize=(15, 9), title="Cases per day, 7 day moving average, with lockdown dates")
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+# uk_projection.deaths_m7.plot(ax=ax)
+for c in COUNTRIES_CORE:
+ lvi = cases_m7[c].last_valid_index()
+# if c != 'UK':
+ ax.text(x = lvi + 1, y = cases_m7[c][lvi], s = f"{c}: {cases_m7[c][lvi]:.0f}")
+ if c in lockdown_dates:
+ for phase in lockdown_dates[c]:
+ marker_col = [l for l in ax.lines if l.get_label() == c][0].get_color()
+ if 'start' in phase:
+ marker_shape = '^'
+ else:
+ marker_shape = 'v'
+ ax.plot(lockdown_dates[c][phase]['since_threshold'], lockdown_dates[c][phase]['cases_m7'],
+ marker_shape,
+ markersize=18, markerfacecolor=marker_col, markeredgecolor=marker_col)
+ if 'start' not in phase:
+ ax.text(x = lockdown_dates[c][phase]['since_threshold'] + 1, y = lockdown_dates[c][phase]['cases_m7'],
+ s = f"{lockdown_dates[c][phase]['date']}: {lockdown_dates[c][phase]['cases_m7']:.0f}")
+plt.savefig('cases_per_day_with_lockdown.png')
+```
+
+```python
+ax = cases_m7[COUNTRIES_CORE].plot(figsize=(10, 6), title="Cases per day, 7 day moving average")
+for c in COUNTRIES_CORE:
+ lvi = cases_m7[c].last_valid_index()
+ ax.text(x = lvi + 1, y = cases_m7[c][lvi], s = c)
+plt.savefig('covid_cases_per_day-core.png')
+```
+
+```python
+ax = deaths_m7[COUNTRIES_CORE].plot(figsize=(15, 9), title="Deaths per day, 7 day moving average",
+ xlim=(0, 15),
+ ylim=(0, 66)
+ )
+# uk_projection.deaths_m7.plot(ax=ax)
+for c in COUNTRIES_CORE:
+ lvi = deaths_m7[c].last_valid_index()
+ if c in lockdown_dates:
+ for phase in lockdown_dates[c]:
+ if 'start' in phase:
+ print(c, phase)
+ marker_col = [l for l in ax.lines if l.get_label() == c][0].get_color()
+ ax.plot(lockdown_dates[c][phase]['since_threshold'], lockdown_dates[c][phase]['deaths_m7'], '*',
+ markersize=18, markerfacecolor=marker_col, markeredgecolor=marker_col)
+ ax.text(x = lockdown_dates[c][phase]['since_threshold'] + 0.3, y = lockdown_dates[c][phase]['deaths_m7'],
+ s = f"{lockdown_dates[c][phase]['date']}: {lockdown_dates[c][phase]['deaths_m7']:.0f}")
+# plt.savefig('school_reopenings.png')
+```
+
+```python
+
+```
+
+```python
+
```
# Write results to summary file
f.write('\n')
f.write('| []() | |\n')
f.write('|:---|---:|\n')
- f.write(f'| Deaths reported so far | {reported_deaths} | \n')
- f.write(f'| Total Covid deaths to date | {uk_covid_deaths * excess_deaths / reported_deaths:.0f} |\n')
+ f.write(f'| Deaths reported so far | {uk_covid_deaths} | \n')
+ f.write(f'| Total Covid deaths to date (estimated) | {uk_deaths_to_date:.0f} |\n')
projection_date = data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['IT']), 'dateRep'].iloc[-1] + pd.Timedelta(s_end - s_start, unit='days')
- f.write(f'| Projected total deaths up to {projection_date.strftime("%Y-%m-%d")} | {uk_projection.deaths.sum() * excess_deaths / reported_deaths:.0f} | \n')
+ f.write(f'| Projected total deaths up to {projection_date.strftime("%Y-%m-%d")} | {deaths_actual_projected_scaled:.0f} | \n')
f.write('\n')
```
f.write('\n')
f.write('### Excess deaths\n')
f.write('\n')
- f.write(f'From week ending 20 March 2020, there have been approximately **{excess_deaths:.0f}** excess deaths, over the average for the previous five years.\n')
+ f.write(f'From week ending 20 March 2020 until week ending {pd.to_datetime(excess_deaths_upto).strftime("%d %B %Y")}, ')
+ f.write(f'there were approximately **{excess_deaths:.0f}** excess deaths, over the average for the previous five years.\n')
f.write('\n')
```
```python
with open('covid_summary.md', 'a') as f:
- f.write(f'In that period, the UK reported {reported_deaths} Covid deaths.\n')
- f.write(f'That means the actual number of Covid death is about {excess_deaths / reported_deaths:.2f} times higher than the reported figures.\n')
+ f.write(f'In that period, the UK reported {ons_reported_deaths} Covid deaths.\n')
+ f.write(f'In the last three weeks for which excess deaths have been reported, the excess deaths have been {excess_death_accuracy:.3f} higher than the Covid-reported deaths.\n')
+# f.write(f'That means the actual number of Covid death is about {excess_deaths / reported_deaths:.2f} times higher than the reported figures.\n')
f.write('\n')
f.write(f'The UK has reported {uk_covid_deaths} deaths so far.\n')
- f.write(f'Using the scaling factor above, I infer that there have been **{uk_covid_deaths * excess_deaths / reported_deaths:.0f}** total deaths so far.\n')
+ f.write(f'Using the scaling factor above (for Covid-19 deaths after the ONS figures), I infer that there have been **{uk_deaths_to_date:.0f}** total deaths so far.\n')
f.write('\n')
```
f.write('\n')
```
+```python
+s_end - s_start - 1
+```
+
```python
with open('covid_summary.md', 'a') as f:
f.write('\n')
f.write('## Projected deaths\n')
f.write(f"The UK's daily deaths data is very similar to Italy's.\n")
- f.write(f'If I use the Italian data for the next {s_end - s_start - 1} days,')
- f.write(f' the UK will report {uk_projection.deaths.sum()} deaths on day {s_end} of the epidemic.\n')
+ f.write(f'If I use the Italian data for the next {s_end - s_start - 1} days (from {s_start_date.strftime("%d %B %Y")} onwards),')
+ f.write(f' the UK will report {uk_projection.deaths.sum()} deaths on day {uk_end} of the epidemic.\n')
f.write('\n')
f.write('Using the excess deaths scaling from above, that will translate into ')
- f.write(f'**{(uk_projection.deaths.sum() * excess_deaths / reported_deaths):.0f}** Covid deaths total.\n')
+ f.write(f'**{deaths_actual_projected_scaled:.0f}** Covid deaths total.\n')
f.write('\n')
```
f.write('\n')
```
+```python
+with open('covid_summary.md', 'a') as f:
+ f.write('\n')
+ f.write('## Cases per day and lockdown dates\n')
+ f.write(f'Based on a 7-day moving average\n')
+ f.write('\n')
+ f.write('\n')
+ f.write('\n')
+```
+
+```python
+with open('covid_summary.md', 'a') as f:
+ f.write('\n')
+ f.write('| Country ID | Country name | Most recent daily cases | Most recent daily deaths |\n')
+ f.write('|:-----------|:-------------|------------------------:|-------------------------:|\n')
+ for c in sorted(COUNTRIES_CORE):
+ lvic = cases_m7[c].last_valid_index()
+ lvid = deaths_m7[c].last_valid_index()
+ f.write(f'| {c} | {countries.loc[c].countriesAndTerritories} | {cases_m7[c][lvic]:.0f} | {deaths_m7[c][lvid]:.0f} | \n')
+ f.write('\n')
+ f.write('(Figures are 7-day averages)\n')
+ f.write('\n')
+```
+
```python
with open('covid_summary.md', 'a') as f:
f.write('# Data sources\n')
* [National Records of Scotland](https://www.nrscotland.gov.uk/statistics-and-data/statistics/statistics-by-theme/vital-events/general-publications/weekly-and-monthly-data-on-births-and-deaths/weekly-data-on-births-and-deaths) (Scotland). Note that Scotland uses ISO8601 week numbers, which start on a Monday.""")
f.write('\n\n')
- f.write('> [Souce code available](https://git.njae.me.uk/?p=covid19.git;a=tree)\n')
+ f.write('> [Source code available](https://git.njae.me.uk/?p=covid19.git;a=tree)\n')
f.write('\n')
```
!scp deaths-radar.png neil@ogedei:/var/www/scripts.njae.me.uk/covid/
!scp covid_deaths_per_day_7.png neil@ogedei:/var/www/scripts.njae.me.uk/covid/
!scp covid_doubling_times_7.png neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp cases_per_day_with_lockdown.png neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+```
+
+```python
+with open('uk_covid_deaths.js', 'w') as f:
+ f.write(f"document.write('{uk_covid_deaths}');")
+
+with open('estimated_total_deaths.js', 'w') as f:
+ f.write(f"document.write('{uk_deaths_to_date:.0f}');")
+
+with open('projection_date.js', 'w') as f:
+ f.write(f"document.write(\'{projection_date.strftime('%d %B %Y')}\');")
+
+with open('projected_deaths.js', 'w') as f:
+ f.write(f"document.write('{uk_projection.deaths.sum():.0f}');")
+
+with open('projected_excess_deaths.js', 'w') as f:
+ f.write(f"document.write('{deaths_actual_projected_scaled:.0f}');")
+
+with open('excess_deaths_upto.js', 'w') as f:
+ f.write(f"document.write('{pd.to_datetime(excess_deaths_upto).strftime('%d %B %Y')}');")
+
+with open('excess_deaths.js', 'w') as f:
+ f.write(f"document.write('{excess_deaths:.0f}');")
+
+with open('reported_deaths.js', 'w') as f:
+ f.write(f"document.write('{ons_reported_deaths:.0f}');")
+
+with open('scaling_factor.js', 'w') as f:
+ f.write(f"document.write('{excess_death_accuracy:.2f}');")
+
+with open('projection_length.js', 'w') as f:
+ f.write(f"document.write('{s_end - s_start - 1}');")
+
+with open('s_end.js', 'w') as f:
+ f.write(f"document.write('{s_end}');")
+
+s_start_date_str = s_start_date.strftime("%d %B %Y")
+with open('s_start_date.js', 'w') as f:
+ f.write(f"document.write('{s_start_date_str}');")
+
+with open('uk_end.js', 'w') as f:
+ f.write(f"document.write('{uk_end}');")
+
+with open('last_uk_date.js', 'w') as f:
+ f.write(f"document.write('{pd.to_datetime(last_uk_date).strftime('%d %B %Y')}');")
+```
+
+```python
+pd.to_datetime(excess_deaths_upto).strftime('%d %B %Y')
+```
+
+```python
+!scp uk_covid_deaths.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp estimated_total_deaths.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp projection_date.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp projected_deaths.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp projected_excess_deaths.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp excess_deaths_upto.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp excess_deaths.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp reported_deaths.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp scaling_factor.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp projection_length.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp s_end.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp s_start_date.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp uk_end.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+!scp last_uk_date.js neil@ogedei:/var/www/scripts.njae.me.uk/covid/
+```
+
+```python
+data_by_date.loc['UK'].to_csv('data_by_day_uk.csv', header=True, index=True)
+```
+
+```python
+ukd = data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['UK']), ['deaths', 'deaths_m7']].droplevel(1)
+ax = ukd.deaths.plot.bar(figsize=(12, 8))
+ukd.deaths_m7.plot.line(ax=ax, color='red')
+# ax = data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['UK']), 'deaths_m7'].plot.line(figsize=(12, 8), color='red')
+# ax = data_since_threshold.replace([np.inf, -np.inf], np.nan).loc[(slice(None), ['UK']), 'deaths'].plot.bar(ax=ax)
+ax.set_xticks(range(0, 120, 20))
+```
+
+```python
+np.arange(0, 130, 20)
+```
+
+```python
+data_by_date.loc['UK']
+```
+
+```python
+data_by_date.loc['UK'].plot(x='deaths_culm', y='deaths', logx=True, logy=True)
+```
+
+```python
+data_by_date.loc['UK'].plot(x='cases_culm', y='cases')
+```
+
+```python
+ukdbd = data_by_date.loc['UK'].copy()
+ukdbd['deaths_m7'] = ukdbd.deaths.transform(lambda x: x.rolling(7, 1).mean())
+ukdbd['cases_m7'] = ukdbd.cases.transform(lambda x: x.rolling(7, 1).mean())
+ukdbd
+```
+
+```python
+ukdbd.plot(x='deaths_culm', y='deaths_m7', logx=True, logy=True)
+```
+
+```python
+fig, ax = plt.subplots(figsize=(12, 8))
+xmax = 10
+for c in COUNTRIES_CORE:
+ if data_since_threshold.loc[(slice(None), c), 'deaths_culm'].max() > xmax:
+ xmax = data_since_threshold.loc[(slice(None), c), 'deaths_culm'].max()
+ data_since_threshold.loc[(slice(None), c), :].plot(x='deaths_culm', y='deaths_m7', logx=True, logy=True, xlim=(10, xmax * 1.1), label=c, ax=ax)
+```
+
+```python
+data_since_threshold.loc[(slice(None), 'UK'), 'deaths_culm'].max()
+```
+
+```python
+countries.continentExp.unique()
+```
+
+```python
+countries.loc['KW']
+```
+
+```python
+data_by_date.groupby(level=0)['deaths'].shift(-25)
+```
+
+```python
+offset_data = data_by_date.loc[:, ['cases']]
+offset_data['deaths'] = data_by_date.groupby(level=0)['deaths'].shift(-25)
+offset_data['cases_m7'] = offset_data.groupby(level=0)['cases'].transform(lambda x: x.rolling(7, 1).mean())
+offset_data['deaths_m7'] = offset_data['deaths'].dropna().groupby(level=0).transform(lambda x: x.rolling(7, 1).mean())
+offset_data['deaths_per_case'] = offset_data.deaths_m7 / offset_data.cases_m7
+offset_data
+```
+
+```python
+deaths_m7
+```
+
+```python
+offset_deaths_m7 = (offset_data.loc[COUNTRIES_ALL, ['deaths_m7']]
+ .unstack().sort_index().xs('deaths_m7', axis=1, drop_level=True)).T.sort_index()
+offset_deaths_m7
+```
+
+```python
+offset_deaths_m7['UK']
+```
+
+```python
+data_since_threshold.loc[(slice(None), 'UK'), :].tail()
+```
+
+```python
+countries.loc['PT']
+```
+
+```python
+ax = cases_m7.iloc[-50:][COUNTRIES_FRIENDS].plot(figsize=(15, 9), title="Cases per day, 7 day moving average", ylim=(-10, 1500))
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+# uk_projection.deaths_m7.plot(ax=ax)
+for c in COUNTRIES_FRIENDS:
+ lvi = cases_m7[c].last_valid_index()
+ if c != 'ES':
+ ax.text(x = lvi + 1, y = cases_m7[c][lvi], s = f"{c}: {cases_m7[c][lvi]:.0f}")
+
+```
+
+```python
+ax = deaths_m7.iloc[-50:][COUNTRIES_FRIENDS].plot(figsize=(15, 9), title="Cases per day, 7 day moving average", ylim=(-10, 100))
+ax.set_xlabel(f"Days since {DEATH_COUNT_THRESHOLD} deaths")
+# uk_projection.deaths_m7.plot(ax=ax)
+for c in COUNTRIES_FRIENDS:
+ lvi = deaths_m7[c].last_valid_index()
+# if c != 'ES':
+ ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = f"{c}: {deaths_m7[c][lvi]:.0f}")
+
```
```python
projects / covid19.git / blobdiff