covid.md

   1 ---
   2 jupyter:
   3   jupytext:
   4     formats: ipynb,md
   5     text_representation:
   6       extension: .md
   7       format_name: markdown
   8       format_version: '1.2'
   9       jupytext_version: 1.3.4
  10   kernelspec:
  11     display_name: Python 3
  12     language: python
  13     name: python3
  14 ---
  15
  16 Data from [European Centre for Disease Prevention and Control](https://www.ecdc.europa.eu/en/publications-data/download-todays-data-geographic-distribution-covid-19-cases-worldwide)
  17
  18 ```python
  19 import itertools
  20 import collections
  21 import pandas as pd
  22 import numpy as np
  23 from scipy.stats import gmean
  24
  25 import matplotlib as mpl
  26 import matplotlib.pyplot as plt
  27 %matplotlib inline
  28 ```
  29
  30 ```python
  31 DEATH_COUNT_THRESHOLD = 10
  32 COUNTRIES_CORE = 'IT DE UK ES IE FR'.split()
  33 COUNTRIES_NORDIC = 'SE NO DK FI UK'.split()
  34 COUNTRIES_FRIENDS = 'IT UK ES BE SI MX'.split()
  35 COUNTRIES_OF_INTEREST = list(set(COUNTRIES_CORE + COUNTRIES_FRIENDS))
  36 COUNTRIES_ALL = list(set(COUNTRIES_CORE + COUNTRIES_FRIENDS + COUNTRIES_NORDIC))
  37 ```
  38
  39 ```python
  40 !curl https://opendata.ecdc.europa.eu/covid19/casedistribution/csv/ > covid.csv
  41 ```
  42
  43 ```python
  44 # First col is a date, treat geoId of NA as 'Namibia', not "NA" value
  45 raw_data = pd.read_csv('covid.csv', parse_dates=[0], keep_default_na=False, dayfirst=True)
  46 ```
  47
  48 ```python
  49 raw_data.size
  50 ```
  51
  52 ```python
  53 raw_data.head()
  54 ```
  55
  56 ```python
  57 raw_data.dtypes
  58 ```
  59
  60 ```python
  61 base_data = raw_data.set_index(['geoId', 'dateRep'])
  62 base_data.sort_index(inplace=True)
  63 base_data
  64 ```
  65
  66 ```python
  67 base_data.loc['UK']
  68 ```
  69
  70 ```python
  71 countries = raw_data[['geoId', 'countriesAndTerritories', 'popData2018']]
  72 countries = countries[countries['popData2018'] != '']
  73 countries = countries.drop_duplicates()
  74 countries.set_index('geoId', inplace=True)
  75 countries = countries.astype({'popData2018': 'int64'})
  76 countries.head()
  77 ```
  78
  79 ```python
  80 countries.shape
  81 ```
  82
  83 ```python
  84 countries[countries.countriesAndTerritories == 'Finland']
  85 ```
  86
  87 ```python
  88 countries.loc[COUNTRIES_OF_INTEREST]
  89 ```
  90
  91 ```python
  92 data_by_date = base_data[['cases', 'deaths']]
  93 data_by_date.head()
  94 ```
  95
  96 ```python
  97 data_by_date.loc['UK']
  98 ```
  99
 100 ```python
 101 data_by_date.groupby(level=0).cumsum()
 102 ```
 103
 104 ```python
 105 data_by_date = data_by_date.merge(
 106     data_by_date.groupby(level=0).cumsum(),
 107     suffixes=('', '_culm'),
 108     left_index=True, right_index=True)
 109 data_by_date
 110 ```
 111
 112 ```python
 113 # data_by_date[data_by_date.deaths_culm > DEATH_COUNT_THRESHOLD]
 114 ```
 115
 116 ```python
 117 # days_since_threshold = data_by_date[data_by_date.deaths_culm > DEATH_COUNT_THRESHOLD].groupby(level=0).cumcount()
 118 # days_since_threshold.rename('since_threshold', inplace=True)
 119 ```
 120
 121 ```python
 122 dbd = data_by_date[data_by_date.deaths_culm > DEATH_COUNT_THRESHOLD].reset_index(level=1)
 123 dbd['since_threshold'] = dbd.dateRep
 124 dbd.set_index('dateRep', append=True, inplace=True)
 125 dbd.sort_index(inplace=True)
 126 days_since_threshold = dbd.groupby(level=0).diff().since_threshold.dt.days.fillna(0).astype(int).groupby(level=0).cumsum()
 127 # days_since_threshold.groupby(level=0).cumsum()
 128
 129 # days_since_threshold = dbd.rename('since_threshold')
 130 days_since_threshold
 131 ```
 132
 133 ```python
 134 # days_since_threshold = (data_by_date[data_by_date.deaths_culm > DEATH_COUNT_THRESHOLD]
 135 #                         .reset_index(level=1).groupby(level=0)
 136 #                         .diff().dateRep.dt.days
 137 #                         .groupby(level=0).cumcount()
 138 #                        )
 139 # days_since_threshold.rename('since_threshold', inplace=True)
 140 # days_since_threshold
 141 ```
 142
 143 ```python
 144 data_since_threshold = data_by_date.merge(days_since_threshold,
 145     left_index=True, right_index=True)
 146 data_since_threshold
 147 ```
 148
 149 ```python
 150 data_since_threshold = data_since_threshold.set_index('since_threshold', append=True
 151                               ).reorder_levels(['since_threshold', 'geoId', 'dateRep']
 152                                               ).reset_index('dateRep')
 153 data_since_threshold
 154 ```
 155
 156 ```python
 157 data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), :]
 158 ```
 159
 160 ```python
 161 data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), ['deaths_culm']].unstack().plot(logy=True)
 162 ```
 163
 164 ```python
 165 deaths = data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT', 'IE']), ['deaths_culm']].unstack().xs('deaths_culm', axis=1, drop_level=True)
 166 ```
 167
 168 ```python
 169 deaths = data_since_threshold.loc[(slice(None), COUNTRIES_ALL), ['deaths_culm']].unstack().xs('deaths_culm', axis=1, drop_level=True)
 170 ```
 171
 172 ```python
 173 data_since_threshold.reset_index().merge(countries, on='geoId').set_index(['since_threshold', 'geoId'])
 174 ```
 175
 176 ```python
 177 data_since_threshold_per_capita = data_since_threshold.reset_index().merge(countries, on='geoId').set_index(['since_threshold', 'geoId'])
 178 data_since_threshold_per_capita['cases_culm_pc'] = data_since_threshold_per_capita.cases_culm / data_since_threshold_per_capita.popData2018
 179 data_since_threshold_per_capita['deaths_culm_pc'] = data_since_threshold_per_capita.deaths_culm / data_since_threshold_per_capita.popData2018
 180 data_since_threshold_per_capita
 181 ```
 182
 183 ```python
 184 deaths_pc = data_since_threshold_per_capita.loc[(slice(None), ['UK', 'DE', 'IT', 'IE']), ['deaths_culm_pc']].unstack().xs('deaths_culm_pc', axis=1, drop_level=True)
 185 ```
 186
 187 ```python
 188 deaths_pc
 189 ```
 190
 191 ```python
 192 deaths_pc = data_since_threshold_per_capita.loc[(slice(None), COUNTRIES_ALL), ['deaths_culm_pc']].unstack().xs('deaths_culm_pc', axis=1, drop_level=True)
 193 ```
 194
 195 ```python
 196 deaths[COUNTRIES_CORE].plot()
 197 ```
 198
 199 ```python
 200 deaths[COUNTRIES_FRIENDS].plot()
 201 ```
 202
 203 ```python
 204 ax = deaths[COUNTRIES_CORE].plot(figsize=(10, 6), title="Total deaths, linear")
 205 for c in COUNTRIES_CORE:
 206     lvi = deaths[c].last_valid_index()
 207     ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
 208 # plt.savefig('covid_deaths_total_linear.png')
 209 ```
 210
 211 ```python
 212 ax = deaths[COUNTRIES_NORDIC].plot(figsize=(10, 6), title="Total deaths, linear")
 213 for c in COUNTRIES_NORDIC:
 214     lvi = deaths[c].last_valid_index()
 215     ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
 216 # plt.savefig('covid_deaths_total_linear.png')
 217 ```
 218
 219 ```python
 220 ax = deaths[COUNTRIES_OF_INTEREST].plot(figsize=(10, 6), title="Total deaths, linear")
 221 for c in COUNTRIES_OF_INTEREST:
 222     lvi = deaths[c].last_valid_index()
 223     ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
 224 plt.savefig('covid_deaths_total_linear.png')
 225 ```
 226
 227 ```python
 228 ax = deaths[COUNTRIES_CORE].plot(logy=True, figsize=(10, 6), title="Total deaths, log")
 229 for c in COUNTRIES_CORE:
 230     lvi = deaths[c].last_valid_index()
 231     ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
 232
 233 plt.savefig('covid_deaths_total_log.png')
 234 ```
 235
 236 ```python
 237 ax = deaths[COUNTRIES_FRIENDS].plot(logy=True, figsize=(10, 6), title="Total deaths, log")
 238 for c in COUNTRIES_FRIENDS:
 239     lvi = deaths[c].last_valid_index()
 240     ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
 241
 242 # plt.savefig('covid_deaths_total_log.png')
 243 ```
 244
 245 ```python
 246 ax = deaths[COUNTRIES_NORDIC].plot(logy=True, figsize=(10, 6), title="Total deaths, log")
 247 for c in COUNTRIES_NORDIC:
 248     lvi = deaths[c].last_valid_index()
 249     ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
 250
 251 # plt.savefig('covid_deaths_total_log.png')
 252 ```
 253
 254 ```python
 255 ax = deaths[COUNTRIES_OF_INTEREST].plot(logy=True, figsize=(10, 6), title="Total deaths, log")
 256 for c in COUNTRIES_OF_INTEREST:
 257     lvi = deaths[c].last_valid_index()
 258     ax.text(x = lvi + 1, y = deaths[c][lvi], s = c)
 259
 260 plt.savefig('covid_deaths_total_log.png')
 261 ```
 262
 263 ```python
 264 deaths_pc.plot(figsize=(10, 6), title="Deaths per capita, linear")
 265 plt.savefig('covid_deaths_per_capita_linear.png')
 266 ```
 267
 268 ```python
 269 ax = deaths_pc.plot(logy=True, figsize=(10, 6), title="Deaths per capita, log")
 270 for c in deaths_pc.columns:
 271     lvi = deaths_pc[c].last_valid_index()
 272     ax.text(x = lvi + 1, y = deaths_pc[c][lvi], s = c)
 273 ```
 274
 275 ```python
 276 deaths_pc[['UK', 'IE']].plot( figsize=(10, 6), title="Deaths per capita, linear")
 277 ```
 278
 279 ```python
 280 deaths_pc[['UK', 'IE']].plot(logy=True, figsize=(10, 6), title="Deaths per capita, log")
 281 ```
 282
 283 ```python
 284 deaths[['UK', 'ES', 'IT']].plot(logy=True, figsize=(10, 6), title="Deaths, log")
 285 plt.savefig('covid_deaths_selected_log.png')
 286 ```
 287
 288 ```python
 289 deaths[['UK', 'ES', 'IT', 'MX']].plot(logy=True, figsize=(10, 6), title="Deaths, log")
 290 ```
 291
 292 ```python
 293 data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), :]
 294 ```
 295
 296 ```python
 297 data_since_threshold['deaths_m4'] = data_since_threshold.groupby(level=1)['deaths'].transform(lambda x: x.rolling(4, 1).mean())
 298 data_since_threshold['deaths_m7'] = data_since_threshold.groupby(level=1)['deaths'].transform(lambda x: x.rolling(7, 1).mean())
 299 data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), :]
 300 ```
 301
 302 ```python
 303 deaths_m4 = (data_since_threshold.loc[(slice(None), COUNTRIES_ALL), ['deaths_m4']]
 304              .unstack().xs('deaths_m4', axis=1, drop_level=True))
 305 ```
 306
 307 ```python
 308 deaths_m7 = (data_since_threshold.loc[(slice(None), COUNTRIES_ALL), ['deaths_m7']]
 309              .unstack().xs('deaths_m7', axis=1, drop_level=True))
 310 ```
 311
 312 ```python
 313 ax = deaths_m4.plot(figsize=(10, 6), title="Deaths per day, 4 day moving average")
 314 for c in deaths_m4.columns:
 315     lvi = deaths_m4[c].last_valid_index()
 316     ax.text(x = lvi + 1, y = deaths_m4[c][lvi], s = c)
 317 plt.savefig('covid_deaths_per_day.png')
 318 ```
 319
 320 ```python
 321 ax = deaths_m4[COUNTRIES_CORE].plot(figsize=(10, 6), title="Deaths per day, 4 day moving average")
 322 for c in COUNTRIES_CORE:
 323     lvi = deaths_m4[c].last_valid_index()
 324     ax.text(x = lvi + 1, y = deaths_m4[c][lvi], s = c)
 325 plt.savefig('covid_deaths_per_day-core.png')
 326 ```
 327
 328 ```python
 329 ax = deaths_m4[COUNTRIES_FRIENDS].plot(figsize=(10, 6), title="Deaths per day, 4 day moving average")
 330 for c in COUNTRIES_FRIENDS:
 331     lvi = deaths_m4[c].last_valid_index()
 332     ax.text(x = lvi + 1, y = deaths_m4[c][lvi], s = c)
 333 plt.savefig('covid_deaths_per_day-friends.png')
 334 ```
 335
 336 ```python
 337 C7s = 'ES FR IT UK'.split()
 338 ax = deaths_m7[C7s].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average")
 339 for c in C7s:
 340     lvi = deaths_m7[c].last_valid_index()
 341     ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = c)
 342 # plt.savefig('covid_deaths_per_day-friends.png')
 343 ```
 344
 345 ```python
 346 ax = deaths_m7[COUNTRIES_FRIENDS].plot(figsize=(10, 6), title="Deaths per day, 7 day moving average")
 347 for c in COUNTRIES_FRIENDS:
 348     lvi = deaths_m7[c].last_valid_index()
 349     ax.text(x = lvi + 1, y = deaths_m7[c][lvi], s = c)
 350 # plt.savefig('covid_deaths_per_day-friends.png')
 351 ```
 352
 353 ```python
 354 def gmean_scale(items):
 355     return gmean(items) / items[-1]
 356 ```
 357
 358 ```python
 359 def doubling_time(df):
 360     return np.log(2) / np.log((df.deaths_culm + df.deaths_g4) / df.deaths_culm)
 361 ```
 362
 363 ```python
 364 # data_since_threshold['deaths_g4'] = data_since_threshold.groupby(level=1)['deaths'].transform(lambda x: x.rolling(4, 1).apply(gmean_scale, raw=True))
 365 # data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), :]
 366 ```
 367
 368 ```python
 369 data_since_threshold['deaths_g4'] = data_since_threshold.groupby(level=1)['deaths'].transform(lambda x: x.rolling(4, 1).apply(gmean, raw=True))
 370 data_since_threshold.loc[(slice(None), ['UK', 'DE', 'IT']), :]
 371 ```
 372
 373 ```python
 374 data_since_threshold['doubling_time'] = data_since_threshold.groupby(level=1).apply(doubling_time).reset_index(level=0, drop=True)
 375 # data_since_threshold.loc[(slice(None), 'UK'), :]
 376 ```
 377
 378 ```python
 379 doubling_times = (data_since_threshold.loc[(slice(None), COUNTRIES_OF_INTEREST), ['doubling_time']]
 380              .unstack().xs('doubling_time', axis=1, drop_level=True))
 381 doubling_times.replace([np.inf, -np.inf], np.nan, inplace=True)
 382 ```
 383
 384 ```python
 385 ax = doubling_times.plot(figsize=(10, 6), title="Doubling times")
 386 for c in doubling_times.columns:
 387     lvi = doubling_times[c].last_valid_index()
 388     ax.text(x = lvi + 1, y = doubling_times[c][lvi], s = c)
 389 # plt.savefig('covid_deaths_per_day.png')
 390 ```
 391
 392 ```python
 393 ax = doubling_times[COUNTRIES_CORE].plot(figsize=(10, 6), title="Doubling times")
 394 for c in COUNTRIES_CORE:
 395     lvi = doubling_times[c].last_valid_index()
 396     ax.text(x = lvi + 1, y = doubling_times[c][lvi], s = c)
 397 plt.savefig('covid_doubling_times.png')
 398 ```
 399
 400 ```python
 401 ax = doubling_times[COUNTRIES_FRIENDS].plot(figsize=(10, 6), title="Doubling times")
 402 for c in COUNTRIES_FRIENDS:
 403     lvi = doubling_times[c].last_valid_index()
 404     ax.text(x = lvi + 1, y = doubling_times[c][lvi], s = c)
 405 plt.savefig('covid_doubling_times_friends.png')
 406 ```
 407
 408 ```python
 409 data_since_threshold.replace([np.inf, -np.inf], np.nan).groupby(level=1).last().loc[COUNTRIES_ALL]#, [doubling_time]]
 410 ```
 411
 412 ```python
 413
 414 ```