-
Notifications
You must be signed in to change notification settings - Fork 0
/
data_loading.py
413 lines (375 loc) · 14.9 KB
/
data_loading.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
#%%
from datetime import datetime
from pathlib import Path
import numpy as np
import pandas as pd
from utility import fix_multilevel
D0 = "0D"
D1 = "1D"
D2 = "2D"
from enum import Enum
class Sources(Enum):
Vaccinations = 1
Deaths = 2
DeathsByAge = 3
Demographics = 4
VaccAt = 5
DeathsAtAge = 6
VaccDe = 7
def get_risk_by_age():
# data from CDC
# https://www.cdc.gov/coronavirus/2019-ncov/images/need-extra-precautions/319360-A_COVID-19_RiskForSevereDisease_Race_Age_2.18_p1.jpg
# tuples denote inclusive ranges
age_risk_factor = {
(0, 4): 2, # X
(5, 17): 1, # reference group A
(18, 29): 15, # B
(30, 39): 45, # C
(40, 49): 130, # D
(50, 64): 400, # E
(65, 74): 1100, # F
(75, 84): 2800, # G
(85, 99): 7900, # H
}
DATA = {
# OWID
Sources.Vaccinations: r"https://raw.githubusercontent.com/owid/covid-19-data/master/public/data/vaccinations/vaccinations.csv",
# OWID, no idea from where
Sources.Deaths: r"https://covid.ourworldindata.org/data/owid-covid-data.csv",
# CDC https://data.cdc.gov/NCHS/Provisional-COVID-19-Death-Counts-by-Sex-Age-and-S/9bhg-hcku/
Sources.DeathsByAge: Path(r"./data/death_count_by_age.csv"),
# OWID, based on UN data
Sources.Demographics: Path(r"./data/owid-population.csv"),
Sources.VaccAt: r"https://info.gesundheitsministerium.at/data/timeline-eimpfpass.csv",
# https://www.ages.at/themen/krankheitserreger/coronavirus/
Sources.DeathsAtAge: Path(r"./data/deaths_by_age_at.csv"),
# https://www.rki.de/DE/Content/InfAZ/N/Neuartiges_Coronavirus/Daten/Impfquoten-Tab.html
Sources.VaccDe: Path(r"./data/de_data.csv"),
}
def get_data_with_cache(name, **kwargs) -> pd.DataFrame:
source = DATA[name]
if isinstance(source, Path):
return pd.read_csv(source, **kwargs)
elif isinstance(source, str):
today = datetime.now().date().strftime("%Y-%m-%d")
path = Path(f"./cache/{name}_{today}.pqt")
if path.exists():
df_raw = pd.read_parquet(path)
else:
df_raw = pd.read_csv(source, **kwargs)
df_raw.to_parquet(path, compression="gzip")
return df_raw
else:
raise Exception("Source must be either string or Path")
def get_vaccination_data(extend_by_days=0, per_million=True) -> pd.DataFrame:
"""Load data from CSV and prepare it for use
Args:
extend_by_days (int, optional): Add days at end of DF, must be >= 0
"""
df_raw = get_data_with_cache(Sources.Vaccinations)
df_raw["date"] = pd.to_datetime(df_raw["date"])
vac_pp_col = (
"daily_vaccinations_per_million" if per_million else "daily_vaccinations"
)
df = df_raw.pivot(index="date", columns="location", values=vac_pp_col)
df = df.fillna(0)
# Add additional rows on bottom
start = df.index.max()
new_range = pd.date_range(start, periods=extend_by_days + 1)[1:]
add_below = pd.DataFrame(0, index=new_range, columns=df.columns)
df = pd.concat([df, add_below])
return df
def get_death_data(extend_by_days=0, smoothed=True) -> pd.DataFrame:
# Smoothed data should lessen artifacts from data-reporting delays
death_column = "new_deaths_smoothed" if smoothed else "new_deaths"
df_raw = get_data_with_cache(Sources.Deaths)
df_raw["date"] = pd.to_datetime(df_raw["date"])
df = df_raw.pivot(index="date", columns="location", values=death_column)
df = df.fillna(0)
# Add additional rows on bottom
start = df.index.max()
new_range = pd.date_range(start, periods=extend_by_days + 1)[1:]
add_below = pd.DataFrame(np.nan, index=new_range, columns=df.columns)
df = pd.concat([df, add_below])
df.ffill()
return df
def get_death_distr_by_age_us() -> pd.DataFrame:
df = get_data_with_cache(Sources.DeathsByAge)
# Only get data from 2020 because vaccination changes distribution in 2021
#! This will mess up things when most old people are vaccinated and distribution changes
df = df[df["Group"] == "By Year"]
df = df[df["Year"] == 2020]
df = df[df["State"] == "United States"]
df = df[df["Sex"] == "All Sexes"]
#
df = df.set_index("Age Group")[["COVID-19 Deaths"]].iloc[2:]
df = df.rename({"COVID-19 Deaths": "Deaths"}, axis=1)
# some age groups are overlapping (why???) - remove those
df = df.transpose()[
[
# "0-17 years",
"1-4 years",
"5-14 years",
"15-24 years",
# "18-29 years",
"25-34 years",
# "30-39 years",
"35-44 years",
# "40-49 years",
"45-54 years",
# "50-64 years",
"55-64 years",
"65-74 years",
"75-84 years",
"85 years and over",
]
].rename(
{
"1-4 years": "0-4", # Mislabeled but matches with UN data and likely irrelevant
"5-14 years": "5-14",
"15-24 years": "15-24",
"25-34 years": "25-34",
"35-44 years": "35-44",
"45-54 years": "45-54",
"55-64 years": "55-64",
"65-74 years": "65-74",
"75-84 years": "75-84",
"85 years and over": "85-99",
},
axis=1,
)
df = df.transpose()
df["DeathShare"] = df["Deaths"] / df["Deaths"].sum()
return df
def get_age_data() -> pd.DataFrame:
#%% get raw data from owid (source is UN afaik)
df = get_data_with_cache(Sources.Demographics)
df = df[df.Year == 2020]
# rename columns to be less wordy
rename = {
"Entity": "Location",
"Under 15 years old (UNWPP, 2017)": "0-15",
"Working age (15-64 years old) (UNWPP, 2017)": "15-64",
"65+ years old (UNWPP, 2017)": "65-99",
"Under 5 years old (UNWPP, 2017)": "0-4",
"5-14 years old (UNWPP, 2017)": "5-14",
"15-24 years old (UNWPP, 2017)": "15-24",
"25-64 years old (UNWPP, 2017)": "25-64",
}
df = df.rename(rename, axis=1)
# age brackets don't match with population data, use data from USA to split up groups
# https://www.census.gov/data/tables/time-series/demo/popest/2010s-national-detail.html
# basically, assume that age distribution of USA is broadly representative of whole world
age_factors = { # each letter group sums up to 1
"0-4": {(0, 4): 19576683 / 19576683}, # A
"5-14": { # B
(5, 9): 20195895 / 40994163,
(10, 14): 20798268 / 40994163,
},
"15-24": { # C
(15, 19): 21054570 / 42687510,
(20, 24): 21632940 / 42687510,
},
"25-64": { # D
(25, 29): 23509016 / 170922904,
(30, 34): 22431305 / 170922904,
(35, 39): 21737521 / 170922904,
(40, 44): 19921623 / 170922904,
(45, 49): 20397751 / 170922904,
(50, 54): 20477151 / 170922904,
(55, 59): 21877391 / 170922904,
(60, 64): 20571146 / 170922904,
},
"65-99": { # E
(65, 69): 17455001 / 54058263,
(70, 74): 14028432 / 54058263,
(75, 79): 9652665 / 54058263,
(80, 84): 6317207 / 54058263,
(85, 99): 6604958 / 54058263,
},
}
# rearrange and drop useless stuff
new_groups = {"Location": df["Location"]}
for group, subgroups in age_factors.items():
for (lower, upper), factor in subgroups.items():
new_groups[f"{lower}-{upper}"] = (df[group] * factor).astype(int)
# as death data 10y intervals, recalculate those brackets:
df = pd.DataFrame(new_groups).set_index("Location")
ages_new_brackets = {
"0-4": ["0-4"],
"5-14": ["5-9", "10-14"],
"15-24": ["15-19", "20-24"],
"25-34": ["25-29", "30-34"],
"35-44": ["35-39", "40-44"],
"45-54": ["45-49", "50-54"],
"55-64": ["55-59", "60-64"],
"65-74": ["65-69", "70-74"],
"75-84": ["75-79", "80-84"],
"85-99": ["85-99"],
}
return pd.DataFrame({k: df[v].sum(axis=1) for k, v in ages_new_brackets.items()})
#%%
def get_death_data_by_age() -> pd.DataFrame:
death_distr = get_death_distr_by_age_us().transpose().loc["DeathShare"]
deaths = get_death_data()
age = get_age_data()
age_t = age.transpose()
us_distribution = age_t["United States"] / age_t["United States"].sum()
countries = set(age.index) & set(deaths.columns)
# Mostly small countries are missing, but also:
# - Czechia, because it's sometimes called Czech Republic
# - Serbia ?
# - Kosovo ?
columns = dict()
for country in countries:
age_distr = age_t[country] / age_t[country].sum() # percentage in each bracket
relative_to_us = age_distr / us_distribution # not sure if this is a good idea
x = relative_to_us * death_distr
death_distr_x = x / x.sum()
for group in death_distr_x.index:
columns[(country, group)] = deaths[country] * death_distr_x[group]
columns
return pd.DataFrame(columns)
#%%
def get_vaccinations_de():
df = get_data_with_cache(Sources.VaccDe, parse_dates=True)
df["Date"] = pd.to_datetime(df["Date"].str.slice(0, 10))
df = df.set_index("Date", drop=True)
# De doesn't publish data by age group
columns = {
D1: ["25-34", "35-44", "45-54", "55-64", "65-74", "75-84", "85-99"],
D2: ["25-34", "35-44", "45-54", "55-64", "65-74", "75-84", "85-99"],
}
cs = []
for k, vs in columns.items():
for v in vs:
cs.append((k, v))
df[(k, v)] = 0
df[(k, vs[-1])] = df[k]
df = fix_multilevel(df[cs])
return df[cs]
get_vaccinations_de()
#%%
#%%
def get_vaccinations_at(filter_region=None):
clean_special_chars = lambda x: x.replace("\ufeff", "")
df = get_data_with_cache(Sources.VaccAt, sep=";", parse_dates=True)
df.columns = map(clean_special_chars, df.columns)
# Remove BOM from first column
# Translate columns
translate = {
"Datum": "Date",
"Bevölkerung": "Population",
"BundeslandID": "StateId",
"EingetrageneImpfungen": "Vacc",
# "EingetrageneImpfungenAstraZeneca_1": "",
# "EingetrageneImpfungenAstraZeneca_2": "",
# "EingetrageneImpfungenBioNTechPfizer_1": "",
# "EingetrageneImpfungenBioNTechPfizer_2": "",
# "EingetrageneImpfungenModerna_1": "",
# "EingetrageneImpfungenModerna_2": "",
"EingetrageneImpfungenPro100": "VaccPer100",
"Name": "Region",
"Teilgeimpfte": "Vacc_1",
"Vollimmunisierte": "Vacc_2",
"TeilgeimpftePro100": "Vacc_1Per100",
"VollimmunisiertePro100": "Vacc_2Per100",
}
df = df.rename(translate, axis=1)
df["Date"] = pd.to_datetime(df["Date"].str.slice(0, 10))
relevant = [
"Date",
"Population",
"Vacc",
"VaccPer100",
"Vacc_1",
"Vacc_2",
"Vacc_1Per100",
"Vacc_2Per100",
]
df_full = df.copy()
regions = dict()
for region in df.Region.unique():
df = df_full[df_full.Region == region]
groups = {
D1: {
"00-24": ["Gruppe_<25_M_1", "Gruppe_<25_W_1", "Gruppe_<25_D_1"],
"25-34": ["Gruppe_25-34_M_1", "Gruppe_25-34_W_1", "Gruppe_25-34_D_1"],
"35-44": ["Gruppe_35-44_M_1", "Gruppe_35-44_W_1", "Gruppe_35-44_D_1"],
"45-54": ["Gruppe_45-54_M_1", "Gruppe_45-54_W_1", "Gruppe_45-54_D_1"],
"55-64": ["Gruppe_55-64_M_1", "Gruppe_55-64_W_1", "Gruppe_55-64_D_1"],
"65-74": ["Gruppe_65-74_M_1", "Gruppe_65-74_W_1", "Gruppe_65-74_D_1"],
"75-84": ["Gruppe_75-84_M_1", "Gruppe_75-84_W_1", "Gruppe_75-84_D_1"],
"85-99": ["Gruppe_>84_M_1", "Gruppe_>84_W_1", "Gruppe_>84_D_1"],
},
D2: {
"00-24": ["Gruppe_<25_M_2", "Gruppe_<25_W_2", "Gruppe_<25_D_2"],
"25-34": ["Gruppe_25-34_M_2", "Gruppe_25-34_W_2", "Gruppe_25-34_D_2"],
"35-44": ["Gruppe_35-44_M_2", "Gruppe_35-44_W_2", "Gruppe_35-44_D_2"],
"45-54": ["Gruppe_45-54_M_2", "Gruppe_45-54_W_2", "Gruppe_45-54_D_2"],
"55-64": ["Gruppe_55-64_M_2", "Gruppe_55-64_W_2", "Gruppe_55-64_D_2"],
"65-74": ["Gruppe_65-74_M_2", "Gruppe_65-74_W_2", "Gruppe_65-74_D_2"],
"75-84": ["Gruppe_75-84_M_2", "Gruppe_75-84_W_2", "Gruppe_75-84_D_2"],
"85-99": ["Gruppe_>84_M_2", "Gruppe_>84_W_2", "Gruppe_>84_D_2"],
},
}
dfn = df[relevant].copy()
# necessary in both
df = df.set_index(["Date"]).sort_index()
dfn = dfn.set_index(["Date"]).sort_index()
for d, g in groups.items():
for k, v in g.items():
# Sum over m/f/x, derive to get daily numbers, fill up empty with 0
dfn[(d, k)] = df[v].sum(axis=1).diff().fillna(0).astype(int)
dfn.columns = [("Meta", x) if isinstance(x, str) else x for x in dfn.columns]
# throw away first row as derivative doesn't really work when multiple regions exist
regions[region] = dfn
for k, v in regions.items():
v.columns = [(k, *x) for x in v.columns]
dft = pd.concat(regions.values(), axis=1)
dft.columns = pd.MultiIndex.from_tuples(dft.columns)
selection = dft[
[
"Österreich",
"Burgenland",
# "KeineZuordnung",
"Kärnten",
"Niederösterreich",
"Oberösterreich",
"Salzburg",
"Steiermark",
"Tirol",
"Vorarlberg",
"Wien",
]
].copy()
return fix_multilevel(selection)
#%%
def get_deaths_by_age_at():
df = get_data_with_cache(Sources.DeathsAtAge)
df = df.set_index("AgeGroup").T
return df
# %%
def get_demographics_at():
df = pd.read_excel("data/population-at.xlsx")
df.columns = [x.strip() for x in df.columns]
print(df.columns)
df = df.rename({"Alter": "Age"}, axis=1)
df["Age"] = [x.split()[0] for x in df["Age"]]
df["Age"] = df["Age"].astype(int)
df = df.set_index("Age")
age_groups = {
"00-24": list(range(00, 24 + 1)),
"25-34": list(range(25, 34 + 1)),
"35-44": list(range(35, 44 + 1)),
"45-54": list(range(45, 54 + 1)),
"55-64": list(range(55, 64 + 1)),
"65-74": list(range(65, 74 + 1)),
"75-84": list(range(75, 84 + 1)),
"85-99": list(range(85, 99 + 1)),
}
new_df = pd.DataFrame(columns=df.columns)
for k, v in age_groups.items():
new_df.loc[k] = df.loc[v].sum()
return new_df
# %%