def death(canton):
#get canton
df = df_import[(df_import["geoRegion"] == canton)]
#make list of all age categories
age_list = df["altersklasse_covid19"].unique()
#get last row of each age category, concat to df_final
df_final = pd.DataFrame([])
for a in age_list:
row = df[df["altersklasse_covid19"] == a].tail(1)[[
"altersklasse_covid19", "sumTotal"
]]
df_final = pd.concat([df_final, row])
df_final.columns = ["Altersklasse", "Todesfälle"]
#export backup to csv
df_final.to_csv("/root/covid_aargau/backups/death/death_{}_{}.csv".format(
canton, backdate(0)),
index=False)
#export to csv
df_final.to_csv(
"/root/covid_aargau/data/death/death_{}.csv".format(canton),
index=False)
def covid_basel(canton):
base_url = "https://raw.githubusercontent.com/openZH/covid_19/master/fallzahlen_kanton_total_csv_v2/COVID19_Fallzahlen_Kanton_{}_total.csv".format(
canton)
df_import = pd.read_csv(base_url)
df = df_import[[
"date", "ncumul_conf", "current_hosp", "current_icu", "ncumul_deceased"
]].copy()
df["date"] = pd.to_datetime(df["date"])
#calculate daily data
df["cases"] = df["ncumul_conf"].diff()
df["new_hosp"] = df["current_hosp"].diff()
df["new_icu"] = df["current_icu"].diff()
df["new_dec"] = df["ncumul_deceased"].diff()
#get last row and same day from the previous week
today = df[df['cases'].notnull()].tail(1)
last_week = df[df["date"] == today.iloc[-1]["date"] - timedelta(days=7)]
#unite both rows in df, rename columns
df2 = pd.concat([today, last_week])
df2.columns = [
"Datum", "Fälle gesamt", "Infizierte im Spital",
"davon auf der Intensiv-Station", "Todesfälle gesamt", "Neue Fälle",
"Veränderung Spital-Belegung", "Veränderung Intensiv-Belegung",
"Todesfälle neu"
]
#reorder columns
col_list = df2.columns.tolist()
myorder = [0, 5, 1, 8, 4, 2, 3, 6, 7]
col_list = [col_list[i] for i in myorder]
df3 = df2[col_list]
#get date for "Zahlen vom..."
date_current_values = df3.iloc[0]["Datum"].date().strftime("%d.%m.%Y")
date_current_values = "Zahlen vom " + date_current_values
#set date as index
df3.set_index("Datum", inplace=True)
#all cols to int
df3 = df3.astype(int)
#transpose
df_final = df3[[
"Neue Fälle", "Fälle gesamt", "Todesfälle neu", "Todesfälle gesamt",
"Infizierte im Spital", "davon auf der Intensiv-Station"
]].T
df_final.columns = [date_current_values, "vor einer Woche"]
#make a backup export of the current data
df_final.to_csv(
"/root/covid_aargau/backups/daily_data_bz/backup_{}_{}.csv".format(
backdate(0), canton))
#export to csv
df_final.to_csv(
"/root/covid_aargau/data/only_AG/daily_data_{}.csv".format(canton))
def test_pos(canton):
dfc = df_import[df_import["geoRegion"] == canton].copy()
dfc["datum"] = pd.to_datetime(dfc["datum"])
dfc2 = dfc[dfc["datum"] >= pd.to_datetime("2020-08-31")]
dfc2.set_index("datum", inplace=True)
dfc3 = dfc2.resample("W")["entries_pos", "entries_neg"].sum()
dfc4 = dfc3[dfc3.index < datetime.today()].copy()
dfc4.index = dfc4.index - timedelta(days=6)
dfc4.columns = ["positiv", "negativ"]
dfc4["Positivitätsrate_{}".format(canton)] = (
(dfc4["positiv"] / (dfc4["positiv"] + dfc4["negativ"])) * 100).round(1)
if canton != "CH":
#make a backup export of the current data
dfc4[["positiv", "negativ"]].T.to_csv(
"/root/covid_aargau/backups/tests/tests_weekly_{}_{}.csv".format(
canton, backdate(0)))
#export to csv
dfc4[["positiv", "negativ"]].T.to_csv(
"/root/covid_aargau/data/tests_weekly/tests_weekly_{}.csv".format(
canton))
if canton == "CH":
global dfch
dfch = pd.concat([dfc4[["Positivitätsrate_{}".format(canton)]], dfch])
else:
dfc_final = dfc4[["Positivitätsrate_{}".format(canton)
]].merge(dfch, left_index=True, right_index=True)
dfc_final.columns = [canton, "Schweiz"]
#make a backup export of the current data
dfc_final.to_csv(
"/root/covid_aargau/backups/positivity/positivity_weekly_{}_{}.csv"
.format(canton, backdate(0)))
#export to csv
dfc_final.to_csv(
"/root/covid_aargau/data/positivity_weekly/positivity_weekly_{}.csv"
.format(canton))
def age_dist_cantons(canton):
base_url = "https://raw.githubusercontent.com/timoll/bag_scrape/master/out/canton_age/"
url_men = base_url + canton + "/" + "men.csv"
url_women = base_url + canton + "/" + "women.csv"
url_total = base_url + canton + "/" + "all.csv"
df_men = pd.read_csv(url_men)
sleep(5)
df_women = pd.read_csv(url_women)
sleep(5)
df_total = pd.read_csv(url_total)
s_men_sum = df_men.loc[:, df_men.columns != "date"].sum()
s_women_sum = df_women.loc[:, df_women.columns != "date"].sum()
s_total_sum = df_total.loc[:, df_total.columns != "date"].sum()
df_men_sum = pd.DataFrame(s_men_sum)
df_men_sum = df_men_sum.reset_index()
df_men_sum = df_men_sum.rename(columns={
"index": "Altersklasse",
0: "Männer"
})
df_women_sum = pd.DataFrame(s_women_sum)
df_women_sum = df_women_sum.reset_index()
df_women_sum = df_women_sum.rename(columns={
"index": "Altersklasse",
0: "Frauen"
})
df_total_sum = pd.DataFrame(s_total_sum)
df_total_sum = df_total_sum.reset_index()
df_total_sum = df_total_sum.rename(columns={
"index": "Altersklasse",
0: "Total"
})
df_gen = df_men_sum.merge(df_women_sum, on="Altersklasse")
df_all = df_gen.merge(df_total_sum, on="Altersklasse")
df_all.drop(df_all.tail(1).index, inplace=True)
#export backup to csv
df_all.to_csv(
"/root/covid_aargau/backups/age/altersverteilung_{}_{}.csv".format(
canton, backdate(0)),
index=False)
#export to csv
df_all.to_csv(
"/root/covid_aargau/data/altersverteilung_{}.csv".format(canton),
index=False)
def antigen(canton):
#get all rows for one canton
df_anti_c = df_import[df_import["geoRegion"] == canton].copy()
#format date, to index
df_anti_c["datum_neu"] = pd.to_datetime(df_anti_c["datum"],
format="%Y-%m-%d")
df_anti_c.set_index("datum_neu", inplace=True)
#new column weekday, select only relevant columns
df_anti_c["weekday"] = df_anti_c.index.weekday
df_anti_c2 = df_anti_c[[
"weekday", "entries", "sumTotal", "sum7d", "nachweismethode"
]].copy()
#get 7-day-value for all sundays from november onwards
df_anti_c2_sun = df_anti_c2[(df_anti_c2["weekday"] == 6) & (
df_anti_c2.index > pd.to_datetime("2020-11"))].copy()
df_anti_final = df_anti_c2_sun[["nachweismethode", "sum7d"]].copy()
df_anti_final["nachweismethode"] = df_anti_final[
"nachweismethode"].str.replace("Antigen_Schnelltest",
"Antigen-Schnelltest")
df_anti_final["nachweismethode"] = df_anti_final[
"nachweismethode"].str.replace("PCR", "PCR (herkömmlich)")
#date minus six days, in order to display monday
df_anti_final.index = df_anti_final.index - timedelta(days=6)
df_anti_final.index = df_anti_final.index.strftime("%d.%m.%Y")
#pivot to perfection
cols = []
for i in df_anti_final.index:
if i not in cols:
cols.append(i)
df_anti_final2 = df_anti_final.pivot_table(index="nachweismethode",
columns=df_anti_final.index,
values="sum7d")
df_anti_final3 = df_anti_final2[cols]
#export backup to csv
df_anti_final3.to_csv(
"/root/covid_aargau/backups/schnelltests/schnelltests_{}_{}.csv".
format(canton, backdate(0)))
#export to csv
df_anti_final3.to_csv(
"/root/covid_aargau/data/schnelltests/schnelltests_{}.csv".format(
canton))
def daily_cases(canton):
#get canton and relevant columns
df = df_import[(df_import["geoRegion"] == canton)][[
"datum", "entries", "mean7d"
]]
df.columns = ["Datum", "Fälle", "7-Tages-Durchschnitt"]
#add a baseline (for visualization purposes in Datawrapper)
df["baseline"] = 0
#export backup to csv
df.to_csv(
"/root/covid_aargau/backups/daily_cases/daily_cases_{}_{}.csv".format(
canton, backdate(0)),
index=False)
#export to csv
df.to_csv("/root/covid_aargau/data/daily_cases/daily_cases_{}.csv".format(
canton),
index=False)
def death_getter(canton):
#get data from github
df = pd.read_csv(
"https://raw.githubusercontent.com/daenuprobst/covid19-cases-switzerland/master/covid19_cases_fatalities_switzerland_bag.csv"
)
#formatting
df["date_new"] = pd.to_datetime(df["date"], format="%Y-%m-%d")
df.pop("date")
df.set_index("date_new", inplace=True)
df.rename_axis(None, inplace=True)
#choose canton
if canton == "OCH":
df_ag = df[(df["canton"] == "SG") | (df["canton"] == "TG") |
(df["canton"] == "AR") | (df["canton"] == "AI")]
elif canton == "ZCH":
df_ag = df[(df["canton"] == "LU") | (df["canton"] == "SZ") |
(df["canton"] == "UR") | (df["canton"] == "OW") |
(df["canton"] == "NW") | (df["canton"] == "ZG")]
else:
df_ag = df[df["canton"] == canton]
df_ag2 = df_ag.drop(["canton", "cases"], axis=1)
#groupby age
s = df_ag2.groupby("age_group")["fatalities"].sum()
df_final = pd.DataFrame(s)
df_final.index.name = "Altersklasse"
df_final.columns = ["Todesfälle"]
#export backup to csv
df_final.to_csv(
"/root/covid_aargau/backups/death/todesfaelle_{}_{}.csv".format(
canton, backdate(0)))
#export to csv
df_final.to_csv("/root/covid_aargau/data/deaths_{}.csv".format(canton))
# In[13]:
#get Nachmeldungen Fälle and Todesfälle
url_yest = "https://raw.githubusercontent.com/makwal/covid_aargau/master/backups/daily_data/backup_{}.csv"
#check if data is updated (updated == yesterday's date)
date_in_df = df_final.iloc[1]["date"]
date_in_df_prev = df_final.iloc[0]["date"]
data_updated = date_in_df == date.today() - timedelta(days=1)
#if data has been updated today...
if data_updated:
#... and it's monday...
if date.today().weekday() == 0:
#... get dfe(arly) from 3 days ago
dfe = pd.read_csv(url_yest.format(backdate(3)))
else:
# or from yesterday on any other day than monday
dfe = pd.read_csv(url_yest.format(backdate(1)))
#if data hasn't been updated today...
else:
#... and it's monday...
if date.today().weekday() == 0:
#get dfe(arly) from 4 days ago
dfe = pd.read_csv(url_yest.format(backdate(4)))
else:
#get dfe(arly) from 2 days ago
dfe = pd.read_csv(url_yest.format(backdate(2)))
#"Zahlen vom " day before yesterday
df_travel.columns = df_travel.iloc[1]
df_travel = df_travel.drop([0, 1])
#remove rows with NaN in col header
df_travel = df_travel.loc[:, df_travel.columns.notnull()]
#formatting
df_travel["date"] = pd.to_datetime(df_travel["date"],
errors="coerce").dt.normalize()
df_travel.columns = ["date", "Fälle neu", "aktuell betreut", "Fälle total"]
# In[4]:
#if Monday (weekday == 0), take Friday as latest values
if date.today().weekday() == 0:
df_travel = df_travel[df_travel["date"] < backdate(2)]
else:
df_travel = df_travel[df_travel["date"] < backdate(0)]
# In[5]:
#fill NaN values with previous value
df_travel = df_travel.fillna(method='ffill')
#choose relevant columns
df_travel = df_travel[["date", "aktuell betreut"]]
# In[6]:
#replace "n.d." no data with NaN
df_travel.loc[df_travel["aktuell betreut"] == "n.d.",
# In[4]:
#formatting
df_iso["date"] = pd.to_datetime(df_iso["date"], errors="coerce")
df_iso.columns = ["date", "new_isolated", "total_curr_isolated", "total_isolated",
"new_quar", "total_curr_quar", "total_quar", "NaN"]
#get rid of last row
df_iso.drop(df_iso.tail(1).index,inplace=True)
#if Monday (weekday == 0), take Friday as latest values
if date.today().weekday() == 0:
df_iso = df_iso[df_iso["date"] < backdate(2)]
else:
df_iso = df_iso[df_iso["date"] < backdate(0)]
# In[5]:
#fill NaN values with previous value
df_iso = df_iso.fillna(method='ffill')
#get relevant rows and columns
df_iso_time = df_iso[["date", "total_curr_isolated", "total_curr_quar"]].copy()
#formatting
df_iso_time["date"] = df_iso_time["date"].dt.normalize()
by="total_vaccinations_per_hundred", ascending=False)
#prepare german country names and iso_codes
dfc = df_import_countries[["name", "alpha2", "alpha3"]].copy()
dfc["alpha3"] = dfc["alpha3"].str.upper()
#merge primary dataset with german country names
df3 = df2.merge(dfc, left_on="iso_code", right_on="alpha3")
#formatting for datawrapper visualization purposes
df3["date"] = pd.to_datetime(df3["date"]).dt.strftime("%d.%m.%Y")
df3["name"] = df3["name"].astype(str) + " ^" + df3["date"] + "^"
df3["alpha2"] = ":" + df3["alpha2"] + ": " + df3["name"]
#create df_final
df_final = df3[["alpha2", "total_vaccinations_per_hundred", "vaccines"]].copy()
df_final.columns = [
"Land ^aktualisiert^", "Impfungen pro 100 Menschen", "Impfstoffe"
]
# In[ ]:
#make a backup export of the current data
df_final.to_csv("/root/covid_aargau/backups/vacc_world/backup_{}.csv".format(
backdate(0)),
index=False)
#export to csv
df_final.to_csv("/root/covid_aargau/data/vaccination/vacc_world.csv",
index=False)
def csv_generator(canton):
#read csv from Github
source_url = "https://raw.githubusercontent.com/maekke/bag_data/master/daten_pro_kanton/bag_data_{}.csv".format(canton)
source_url_all = "https://raw.githubusercontent.com/maekke/bag_data/master/bag_weekly_data.csv"
df_testpos = pd.read_csv(source_url)
sleep(2)
df_testpos_all = pd.read_csv(source_url_all)
df_testpos["positive_tests"] = (df_testpos["total_number_of_tests"] / 100 * df_testpos["positivity_rate_percent"]).round(0).astype(int)
df_testpos["negative_tests"] = df_testpos["total_number_of_tests"] - df_testpos["positive_tests"]
#convert calendar week to first day of respective week (date)
def get_weekday(s):
prefix = "2020-W"
cal_week = prefix + str(s)
first_weekday = datetime.strptime(cal_week + '-1', "%G-W%V-%u")
return first_weekday
df_testpos["week_date"] = df_testpos["week"].apply(get_weekday)
df_testpos_all["week_date"] = df_testpos_all["week"].apply(get_weekday)
#formatting
df_testpos["week_date"] = df_testpos["week_date"].dt.strftime("%d.%m.%Y")
df_testpos_all["week_date"] = df_testpos_all["week_date"].dt.strftime("%d.%m.%Y")
df_testpos.columns = ["Kalenderwoche",
"Tests total",
"Positivitätsrate",
"source_file",
"positiv",
"negativ",
"Datum"]
#get relevant columns
df_tests = df_testpos[["Datum", "positiv", "negativ"]].T
df_pos = df_testpos[["Kalenderwoche", "Datum", "Positivitätsrate"]]
df_pos = df_pos.join(df_testpos_all["positivity_rate_percent"])
df_pos.columns = ["Kalenderwoche", "Datum", canton, "Schweiz"]
#make a backup export of the current data
df_tests.to_csv("/root/covid_aargau/backups/tests/tests_weekly_{}_{}.csv".format(canton, backdate(0)))
df_pos.to_csv("/root/covid_aargau/backups/positivity/positivity_weekly_{}_{}.csv".format(canton, backdate(0)))
#export to csv
df_tests.to_csv("/root/covid_aargau/data/tests_weekly_{}.csv".format(canton), header=None)
df_pos.to_csv("/root/covid_aargau/data/positivity_weekly_{}.csv".format(canton), index=False)