def increment_localCasePerCounty(self):
report_date_list = self.getReportDate()
county_list = self.getCounty()
nb_cases_list = self.getNbCases()
## Initialize stock
county_key_list = ['total'] + self.county_key_list
stock = ccm.initializeStock_dailyCounts(county_key_list)
ind_max = 0
## Loop over series
for report_date, county, nb_cases in zip(report_date_list, county_list,
nb_cases_list):
if 'unknown' == county:
continue
ind = ccm.indexForOverall(report_date)
ind_max = max(ind_max, ind + 1)
try:
stock['total'][ind] += nb_cases
stock[county][ind] += nb_cases
except IndexError: ## If NaN
pass
ind_today = ccm.getTodayOrdinal() - ccm.ISODateToOrd(ccm.ISO_DATE_REF)
ind = max(ind_max, ind_today - 1) ## Take the max of data date & today
stock = {k: v[:ind] for k, v in stock.items()}
## Moving average
for key in county_key_list:
stock[key + '_avg'] = ccm.makeMovingAverage(stock[key])
return stock
def increment_incidenceEvolutionByAge(self):
report_date_list = self.getReportDate()
age_list = self.getAge()
nb_cases_list = self.getNbCases()
## Reverse
age_key_list = ['total'] + self.age_key_list[::-1]
## Initialize stock
stock = ccm.initializeStock_dailyCounts(age_key_list)
## Loop over series
for report_date, age, nb_cases in zip(report_date_list, age_list,
nb_cases_list):
ind = ccm.indexForOverall(report_date)
try:
stock['total'][ind] += nb_cases
except IndexError:
pass
else:
try:
stock[age][ind] += nb_cases
except KeyError:
pass
return stock
def saveCsv_incidenceMap(self, mode='both'):
if mode in ['data', 'both']:
stock_dict = self.increment_incidenceMap()
data_l = self.makeLabel_incidenceMap(do_im=True)
data_l = pd.DataFrame(data_l)
## Loop over group
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
stock_col = self.makeLabel_caseByAge(gr)
col_tag_list = stock_col['key']
stock = stock_dict[gr]
data_c = {'county': ['total'] + self.county_key_list}
data_c.update({
col_tag: case_hist.values()
for col_tag, case_hist in zip(col_tag_list, stock)
})
data_c = pd.DataFrame(data_c)
## Save
name = '{}processed_data/{}/incidence_map.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_c)
name = '{}processed_data/{}/incidence_map_label.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_l)
if mode in ['readme', 'both']:
self.makeReadme_incidenceMap(gr)
return
def increment_diffByTransmission(self):
report_date_list = self.getReportDate()
entry_date_list = self.getEntryDate()
onset_date_list = self.getOnsetDate()
trans_list = self.getTransmission()
trans_key_list = ['total'] + self.trans_key_list
stock = {key: [] for key in trans_key_list}
stock_dict = ccm.initializeStockDict_general(stock)
for report_date, entry_date, onset_date, trans in zip(report_date_list, entry_date_list, onset_date_list, trans_list):
if trans != trans:
continue
if trans in ['imported', 'local']:
key = trans
else:
key = 'others'
ord_rep = ccm.ISODateToOrd(report_date)
ord_entry = ccm.ISODateToOrd(entry_date) if entry_date == entry_date else 0
ord_onset = ccm.ISODateToOrd(onset_date) if onset_date == onset_date else 0
diff = min(ord_rep-ord_entry, ord_rep-ord_onset)
index_list = ccm.makeIndexList(report_date)
for ind, stock in zip(index_list, stock_dict.values()):
if ind != ind: ## If NaN
continue
stock['total'].append(diff)
stock[key].append(diff)
return stock_dict
def increment_ageSymptomCorr(self):
report_date_list = self.getReportDate()
trans_list = self.getTransmission()
age_list = self.getAge()
symp_list = self.getSymptom()
stock = {'x_list_list': [], 'y_list_list': [], 'nb_dict': {'N_total': 0, 'N_data': 0}}
stock_dict = ccm.initializeStockDict_general(stock)
## Loop over case
for report_date, trans, age, symp in zip(report_date_list, trans_list, age_list, symp_list):
if trans != trans:
continue
index_list = ccm.makeIndexList(report_date)
for ind, stock in zip(index_list, stock_dict.values()):
if ind != ind: ## If NaN
continue
stock['nb_dict']['N_total'] += 1
## Remove NaN
if age != age or symp != symp:
continue
stock['nb_dict']['N_data'] += 1
stock['x_list_list'].append(symp)
stock['y_list_list'].append(age)
return stock_dict
def setCaseCounts(self):
date_list = self.getDate()
new_soldiers_list = self.getNewSoldiers()
new_impored_list = self.getNewImported()
new_local_list = self.getNewLocal()
new_unclear_list = self.getNewUnclear()
new_others_list = new_soldiers_list + new_unclear_list
new_cases_list = new_impored_list + new_local_list + new_others_list
stock = {'date': date_list[:-1], 'total': new_cases_list[:-1]}
stock = pd.DataFrame(stock)
stock = ccm.adjustDateRange(stock)
cum_cases_dict = {}
for gr in ccm.GROUP_LIST:
data = ccm.truncateStock(stock, gr)
cum_cases = data['total'].values
ind = ~pd.isna(cum_cases)
cum_cases_dict[gr] = cum_cases[ind].sum()
self.n_total = int(cum_cases_dict[ccm.GROUP_OVERALL])
self.n_latest = int(cum_cases_dict[ccm.GROUP_LATEST])
for year in ccm.GROUP_YEAR_LIST:
self.n_year_dict[year] = int(cum_cases_dict[year])
return
def increment_deathDelay(self):
report_date_list = self.getReportDate()
confirmed_date_list = self.getConfirmedDate()
stock = {'list': []}
stock_dict = ccm.initializeStockDict_general(stock)
for report_date, confirmed_date in zip(report_date_list,
confirmed_date_list):
if confirmed_date != confirmed_date:
continue
ord_rep = ccm.ISODateToOrd(report_date)
ord_cf = ccm.ISODateToOrd(confirmed_date)
diff = ord_rep - ord_cf
index_list = ccm.makeIndexList(report_date)
for ind, stock in zip(index_list, stock_dict.values()):
if ind != ind: ## If NaN
continue
stock['list'].append(diff)
return stock_dict
def saveCsv_statusEvolution(self, mode='both'):
if mode in ['data', 'both']:
date_list = self.getDate()
cum_deaths_list = self.getCumDeaths()
cum_dis_list = self.getCumDischarged()
cum_hosp_list = self.getCumHospitalized()
stock = {
'date': date_list,
'discharged': cum_dis_list,
'hospitalized': cum_hosp_list,
'death': cum_deaths_list
}
stock = pd.DataFrame(stock)
stock = ccm.adjustDateRange(stock)
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
data = ccm.truncateStock(stock, gr)
## Save
name = '{}processed_data/{}/status_evolution.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data)
if mode in ['readme', 'both']:
self.makeReadme_statusEvolution(gr)
return
def saveCsv_caseByAge(self, mode='both'):
if mode in ['data', 'both']:
stock_dict = self.increment_caseByAge()
## Loop over group
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
data_l = self.makeLabel_caseByAge(gr)
data_l = pd.DataFrame(data_l)
col_tag_list = data_l['key']
stock = stock_dict[gr]
data_c = {'age': self.age_key_list}
data_c.update({
col_tag: case_hist.values()
for col_tag, case_hist in zip(col_tag_list, stock)
})
data_c = pd.DataFrame(data_c)
## Save
name = '{}processed_data/{}/case_by_age.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_c)
name = '{}processed_data/{}/case_by_age_label.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_l)
if mode in ['readme', 'both']:
self.makeReadme_caseByAge(gr)
return
def saveCsv_deathDelay(self, mode='both'):
if mode in ['data', 'both']:
stock_dict = self.increment_deathDelay()
stock_dict = self.makeHist_deathDelay(stock_dict)
stock_dict_c, stock_dict_l = self.makeStock_deathDelay(stock_dict)
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
data_c = stock_dict_c[gr]
data_c = pd.DataFrame(data_c)
data_l = stock_dict_l[gr]
data_l = pd.DataFrame(data_l)
## Save
name = '{}processed_data/{}/death_delay.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_c)
name = '{}processed_data/{}/death_delay_label.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_l)
if mode in ['readme', 'both']:
self.makeReadme_deathDelay(gr)
return
def increment_incidenceMap(self):
report_date_list = self.getReportDate()
county_list = self.getCounty()
nb_cases_list = self.getNbCases()
## Initialize stock dict
county_key_list = ['total'] + self.county_key_list
case_hist = {key: 0 for key in county_key_list}
stock = [case_hist.copy() for i in range(13)]
stock_dict = ccm.initializeStockDict_general(stock)
## Add empty hist for overall (all year + 12 months per year)
nb_years = len(ccm.GROUP_YEAR_LIST)
for i in range(13 * nb_years - 12):
stock_dict[ccm.GROUP_OVERALL].append(case_hist.copy())
## Loop over series
for report_date, county, nb_cases in zip(report_date_list, county_list,
nb_cases_list):
if 'unknown' == county:
continue
index_list = ccm.makeIndexList(report_date)
for ind, gr, stock in zip(index_list, stock_dict.keys(),
stock_dict.values()):
if ind != ind:
continue
stock[0]['total'] += nb_cases
stock[0][county] += nb_cases
if ccm.GROUP_LATEST == gr:
lookback_week = (
ind - ccm.NB_LOOKBACK_DAYS
) // 7 ## ind - ccm.NB_LOOKBACK_DAYS in [-90, -1]; this will be in [-13, -1]
if lookback_week >= -12:
stock[-lookback_week]['total'] += nb_cases
stock[-lookback_week][county] += nb_cases
elif ccm.GROUP_OVERALL == gr:
y_ind = int(report_date[:4]) - 2020
m_ind = int(report_date[5:7])
yy_ind = 1 + 13 * y_ind
ym_ind = 1 + m_ind + 13 * y_ind
stock[yy_ind]['total'] += nb_cases
stock[yy_ind][county] += nb_cases
stock[ym_ind]['total'] += nb_cases
stock[ym_ind][county] += nb_cases
else:
mm = int(report_date[5:7])
stock[mm]['total'] += nb_cases
stock[mm][county] += nb_cases
return stock_dict
def saveCsv_evtTimeline(self, mode='both'):
if mode in ['data', 'both']:
data = self.makeStock_evtTimeline()
data = pd.DataFrame(data)
name = '{}processed_data/event_timeline_zh-tw.csv'.format(
ccm.DATA_PATH)
ccm.saveCsv(name, data)
if mode in ['readme', 'both']:
self.makeReadme_evtTimeline()
return
def saveCsv_all():
mode = 'both'
ccm.initializeReadme()
print()
case_sheet = COVID_case.CaseSheet()
case_sheet.saveCsv(mode=mode)
print()
status_sheet = COVID_status.StatusSheet()
status_sheet.saveCsv(mode=mode)
print()
test_sheet = COVID_test.TestSheet()
test_sheet.saveCsv(mode=mode)
print()
border_sheet = COVID_border.BorderSheet()
border_sheet.saveCsv(mode=mode)
print()
timeline_sheet = COVID_timeline.TimelineSheet()
timeline_sheet.saveCsv(mode=mode)
print()
county_sheet = COVID_county.CountySheet()
county_sheet.saveCsv(mode=mode)
print()
vacc_sheet = COVID_vaccination.VaccinationSheet()
vacc_sheet.saveCsv(mode=mode)
print()
vc_sheet = COVID_vaccination_county.VaccinationCountySheet()
vc_sheet.saveCsv(mode=mode)
print()
death_sheet = COVID_death.DeathSheet()
death_sheet.saveCsv(mode=mode)
print()
saveCsv_others(status_sheet,
test_sheet,
border_sheet,
county_sheet,
death_sheet,
mode=mode)
print()
ccm.saveMarkdown_readme()
print()
return
def makeStock_caseFatalityRates(self):
date_list = self.getDate()
new_soldiers_list = self.getNewSoldiers()
new_imported_list = self.getNewImported()
new_local_list = self.getNewLocal()
new_unclear_list = self.getNewUnclear()
new_deaths_list = self.getNewDeaths()
cum_deaths_list = self.getCumDeaths()
new_case_list = new_imported_list + new_local_list + new_soldiers_list + new_unclear_list
cum_case_list = np.cumsum(new_case_list)
offset_days = 12
offset_case_list = np.concatenate([[0] * offset_days,
new_case_list[:-offset_days]])
offset_case_list = ccm.sevenDayMovingAverage(offset_case_list)
new_deaths_list = ccm.sevenDayMovingAverage(new_deaths_list)
cum_case_list, _ = ccm.pandasNAToZero(cum_case_list)
cum_deaths_list, _ = ccm.pandasNAToZero(cum_deaths_list)
with wng.catch_warnings(): ## Avoid division by zero
wng.simplefilter('ignore')
front = (offset_case_list != offset_case_list) + (offset_case_list
== 0)
back = front.copy()
front[-7:] = False ## Front blank forced to 0
back[:7] = False ## Back blank forced to NaN
inst_cfr_list = new_deaths_list / offset_case_list
inst_cfr_list = np.around(inst_cfr_list, decimals=6)
inst_cfr_list[front] = 0
inst_cfr_list[back] = np.nan
front = (cum_case_list != cum_case_list) + (cum_case_list == 0)
back = front.copy()
front[-7:] = False ## Front blank forced to 0
back[:7] = False ## Back blank forced to NaN
cum_cfr_list = cum_deaths_list / cum_case_list
cum_cfr_list = np.around(cum_cfr_list, decimals=6)
cum_cfr_list[front] = 0
cum_cfr_list[back] = np.nan
stock = {
'date': date_list,
'inst_CFR': inst_cfr_list,
'cumul_CFR': cum_cfr_list
}
return stock
def saveCsv_caseByDetection(self, mode='both'):
gr = ccm.GROUP_2020
if mode in ['data', 'both']:
stock_r, stock_o = self.increment_caseByDetection()
stock_r = pd.DataFrame(stock_r)
stock_r = ccm.adjustDateRange(stock_r)
stock_o = pd.DataFrame(stock_o)
stock_o = ccm.adjustDateRange(stock_o)
## Make data frame
data_r = ccm.truncateStock(stock_r, gr)
data_o = ccm.truncateStock(stock_o, gr)
## Save
name = '{}processed_data/{}/case_by_detection_by_report_day.csv'.format(ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_r)
name = '{}processed_data/{}/case_by_detection_by_onset_day.csv'.format(ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_o)
if mode in ['readme', 'both']:
self.makeReadme_caseByDetection(gr)
return
def __init__(self, verbose=True):
self.coltag_row_id = 'id'
self.coltag_report_date = '公布日'
self.coltag_id = '案號'
self.coltag_gender = '性別'
self.coltag_age = '年齡'
self.coltag_comorbidity = '慢性病史'
self.coltag_vacc_status = '是否接種COVID19疫苗'
self.coltag_contact_hist = '活動接觸史'
self.coltag_onset_date = '發病日'
self.coltag_symptoms = '症狀'
self.coltag_tested_date = '採檢日'
self.coltag_isolation_date = '住院/隔離日'
self.coltag_confirmed_date = '確診日'
self.coltag_death_date = '死亡日'
name = '{}raw_data/COVID-19_in_Taiwan_raw_data_death.csv'.format(
ccm.DATA_PATH)
data = ccm.loadCsv(name, verbose=verbose)
## https://covid-19.nchc.org.tw/api/[email protected]&querydata=4002
## https://covid-19.nchc.org.tw/api/download/covidtable_taiwan_cdc_death_2022-06-22.csv
self.data = data
self.correctData()
self.n_total = self.getNbRows()
self.age_key_list = [
'0-9', '10-19', '20-29', '30-39', '40-49', '50-59', '60-69',
'70-79', '80-89', '90-99', '100+'
]
if verbose:
print('N_total = {:d}'.format(self.n_total))
return
def __init__(self, verbose=True):
self.coltag_row_id = 'ID'
self.coltag_report_date = '發佈統計日期'
self.coltag_county = '縣市'
self.coltag_age = '群組'
self.coltag_1st_dose = '第1劑'
self.coltag_2nd_dose = '第2劑'
self.coltag_3rd_dose_2 = '加強劑'
self.coltag_3rd_dose_1 = '追加劑'
name = '{}raw_data/COVID-19_in_Taiwan_raw_data_vaccination_county.csv'.format(ccm.DATA_PATH)
data = ccm.loadCsv(name, verbose=verbose)
## https://covid-19.nchc.org.tw/api/[email protected]&querydata=2006
## https://covid-19.nchc.org.tw/api/download/vaccineCityLevel_2022-06-22.csv
self.data = data
self.n_total = len(set(self.getDate()))
self.county_key_list = [
'Keelung', 'Taipei', 'New_Taipei', 'Taoyuan', 'Hsinchu', 'Hsinchu_C', 'Miaoli',
'Taichung', 'Changhua', 'Nantou', 'Yunlin',
'Chiayi' ,'Chiayi_C', 'Tainan', 'Kaohsiung', 'Pingtung',
'Yilan', 'Hualien', 'Taitung',
'Penghu', 'Kinmen', 'Matsu',
]
self.age_key_list = ['6-11', '12-17', '18-29', '30-49', '50-64', '65-74', '75+']
if verbose:
print('N_total = {:d}'.format(self.n_total))
return
def makeStock_testPositiveRate(status_sheet, test_sheet):
stock = {}
status_sheet.updateNewCaseCounts(stock)
test_sheet.updateNewTestCounts(stock)
## Smooth
for key, value_arr in stock.items():
if 'date' == key:
continue
value_arr = ccm.sevenDayMovingAverage(value_arr)
ind = value_arr != value_arr
value_arr[ind] = 0
## Push back
stock[key] = value_arr
stock_new = {}
stock_new['date'] = stock['date']
with wng.catch_warnings(): ## Avoid division by zero
wng.simplefilter('ignore')
ind = stock['new_tests'] == 0
value_arr = stock['new_cases'] / stock['new_tests']
value_arr = np.around(value_arr, decimals=4)
value_arr[ind] = np.nan
stock_new['positivity'] = value_arr
return stock_new
def __init__(self, verbose=True):
self.coltag_date = '日期'
self.coltag_week_nb = '週次'
self.coltag_new_cases = '新增確診'
self.coltag_new_cases_per_week = '每週新增確診'
self.coltag_cum_cases = '確診總人數'
self.coltag_new_males = '新增男性'
self.coltag_cum_males = '男性總數'
self.coltag_male_frac = '確診男性率'
self.coltag_new_females = '新增女性'
self.coltag_cum_females = '女性總數'
self.coltag_female_frac = '確診女性率'
self.coltag_new_soldiers = '新增軍人'
self.coltag_cum_soldiers = '敦睦總人數'
self.coltag_soldier_frac = '軍隊率'
self.coltag_new_imported = '新增境外'
self.coltag_cum_imported = '境外總人數'
self.coltag_imported_frac = '境外率'
self.coltag_new_local = '新增本土'
self.coltag_cum_local = '本土總人數'
self.coltag_local_frac = '本土率'
self.coltag_new_unclear = '新增不明'
self.coltag_cum_unclear = '不明總人數'
self.coltag_unclear_frac = '不明率'
self.coltag_new_deaths = '新增死亡'
self.coltag_cum_deaths = '死亡總人數'
self.coltag_death_frac = '死亡率'
self.coltag_new_unknown = '當日未知感染源數'
self.coltag_cum_unknown = '未知感染源總數'
self.coltag_cum_known = '已知感染源數'
self.coltag_new_dis = '新增解除隔離'
self.coltag_cum_dis = '解除隔離數'
self.coltag_cum_dis_and_deaths = '解除隔離+死亡'
self.coltag_cum_hosp = '未解除隔離數'
self.coltag_notes = '備註'
self.n_total = 0
self.n_latest = 0
self.n_year_dict = {}
self.ind_2021 = None
self.ind_2022 = None
self.ind_2023 = None
## new_year_token
name = '{}raw_data/COVID-19_in_Taiwan_raw_data_status_evolution.csv'.format(
ccm.DATA_PATH)
data = ccm.loadCsv(name, verbose=verbose)
self.setData(data)
self.setCaseCounts()
self.trans_key_list = ['imported', 'local', 'others']
if verbose:
print('N_total = {:d}'.format(self.n_total))
print('N_latest = {:d}'.format(self.n_latest))
for year in ccm.GROUP_YEAR_LIST:
print('N_{} = {:d}'.format(year, self.n_year_dict[year]))
return
def saveCsv_incidenceRates(status_sheet, border_sheet, mode='both'):
if mode in ['data', 'both']:
stock = makeStock_incidenceRates(status_sheet, border_sheet)
stock = pd.DataFrame(stock)
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
data = ccm.truncateStock(stock, gr)
## Save
name = '{}processed_data/{}/incidence_rates.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data)
if mode in ['readme', 'both']:
makeReadme_incidenceRates(gr)
return
def saveCsv_testPositiveRate(status_sheet, test_sheet, mode='both'):
if mode in ['data', 'both']:
stock = makeStock_testPositiveRate(status_sheet, test_sheet)
stock = pd.DataFrame(stock)
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
data = ccm.truncateStock(stock, gr)
## Save
name = '{}processed_data/{}/test_positive_rate.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data)
if mode in ['readme', 'both']:
makeReadme_testPositiveRate(gr)
return
def makeLabel_caseByAge(self, gr):
key_list = ['total']
label_list_en = ['Total']
label_list_fr = ['Totaux']
label_list_zh = ['合計']
if gr == ccm.GROUP_LATEST:
for i in range(12):
str_ = '{:d}-{:d}'.format(7 * i, 7 * i + 6)
key_list.append('week_-{:d}'.format(i + 1))
label_list_en.append('{} days ago'.format(str_))
label_list_fr.append('{} jours plus tôt'.format(str_))
label_list_zh.append('{}天前'.format(str_))
elif gr == ccm.GROUP_OVERALL:
for year in ccm.GROUP_YEAR_LIST:
key_list.append(year)
label_list_en.append('{} all year'.format(year))
label_list_fr.append('Année {}'.format(year))
label_list_zh.append('{}全年'.format(year))
for i in range(12):
str_ = '{:d}-{:d}'.format(7 * i, 7 * i + 6)
key_list.append('{}_{}'.format(ccm.numMonthToAbbr(i + 1),
year))
label_list_en.append('{} {}'.format(
ccm.MONTH_NAME_EN[i], year))
label_list_fr.append('{} {}'.format(
ccm.MONTH_NAME_FR[i], year))
label_list_zh.append('{}年{:d}月'.format(year, i + 1))
elif gr in ccm.GROUP_YEAR_LIST:
for i in range(12):
str_ = '{:d}-{:d}'.format(7 * i, 7 * i + 6)
key_list.append('{}'.format(ccm.numMonthToAbbr(i + 1)))
label_list_en.append('{}'.format(ccm.MONTH_NAME_EN[i]))
label_list_fr.append('{}'.format(ccm.MONTH_NAME_FR[i]))
label_list_zh.append('{:d}月'.format(i + 1))
stock = {
'key': key_list,
'label': label_list_en,
'label_fr': label_list_fr,
'label_zh': label_list_zh
}
return stock
def saveCsv_borderStats(self, mode='both'):
if mode in ['data', 'both']:
stock = self.makeStock_borderStats()
stock = pd.DataFrame(stock)
stock = ccm.adjustDateRange(stock)
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
data = ccm.truncateStock(stock, gr)
## Save
name = '{}processed_data/{}/border_statistics.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data)
if mode in ['readme', 'both']:
self.makeReadme_borderStats(gr)
return
def updateNewEntryCounts(self, stock):
ord_ref = ccm.ISODateToOrd(ccm.ISO_DATE_REF)
ord_today = ccm.getTodayOrdinal()
nb_days = ord_today - ord_ref
stock['new_entries'] = np.zeros(nb_days, dtype=int) + np.nan
date_list = self.getDate()
entry_list = self.getEntry()
for date, entry in zip(date_list, entry_list):
ind = ccm.ISODateToOrd(date) - ord_ref
if ind < 0 or ind >= nb_days:
print('Bad ind_r = {:d}'.format(ind))
continue
stock['new_entries'][ind] = entry
return
def saveCsv_deathCounts(self, mode='both'):
if mode in ['data', 'both']:
stock = self.increment_deathCounts()
stock = pd.DataFrame(stock)
stock = ccm.adjustDateRange(stock)
for gr in ccm.GROUP_LIST:
if mode in ['data', 'both']:
data = ccm.truncateStock(stock, gr)
## Save
name = '{}processed_data/{}/death_counts.csv'.format(
ccm.DATA_PATH, gr)
ccm.saveCsv(name, data)
if mode in ['readme', 'both']:
self.makeReadme_deathCounts(gr)
return
def saveCsv_diffByTransmission(self, mode='both'):
gr = ccm.GROUP_2020
if mode in ['data', 'both']:
stock_dict = self.increment_diffByTransmission()
stock_dict = self.makeHist_diffByTransmission(stock_dict)
col_tag_list = ['difference', 'total'] + self.trans_key_list
stock = stock_dict[gr]
data = {key: stock[key] for key in col_tag_list}
data = pd.DataFrame(data)
name = '{}processed_data/{}/difference_by_transmission.csv'.format(ccm.DATA_PATH, gr)
ccm.saveCsv(name, data)
if mode in ['readme', 'both']:
self.makeReadme_diffByTransmission(gr)
return
def increment_deathCounts(self):
report_date_list = self.getReportDate()
## Initialize stocks
col_tag_list = ['death']
stock = ccm.initializeStock_dailyCounts(col_tag_list)
## Loop over cases
for report_date in report_date_list:
try:
ind = ccm.indexForOverall(report_date)
stock['death'][ind] += 1
except IndexError: ## If NaN
pass
## Make avg
for key in col_tag_list:
stock[key + '_avg'] = ccm.makeMovingAverage(stock[key])
return stock
def increment_caseByTransmission(self):
report_date_list = self.getReportDate()
onset_date_list = self.getOnsetDate()
trans_list = self.getTransmission()
link_list = self.getLink()
## Initialize stocks
trans_key_list = ['imported', 'linked', 'unlinked', 'fleet', 'plane', 'unknown']
stock_r = ccm.initializeStock_dailyCounts(trans_key_list)
stock_o = ccm.initializeStock_dailyCounts(trans_key_list)
## Loop over cases
for report_date, onset_date, trans, link in zip(report_date_list, onset_date_list, trans_list, link_list):
if trans != trans:
continue
## Determine column tag
if trans == 'local':
if link == 'unlinked':
key = link
else:
key = 'linked'
else:
key = trans
try:
ind = ccm.indexForOverall(report_date)
stock_r[key][ind] += 1
except IndexError: ## If NaN
pass
## Check if NaN
if onset_date != onset_date:
continue
try:
ind = ccm.indexForOverall(onset_date)
stock_o[key][ind] += 1
except IndexError: ## If NaN
pass
return stock_r, stock_o
def updateNewTestCounts(self, stock):
date_list = self.getDate()
total_list = self.getTotal()
stock_tmp = {'date': date_list, 'new_tests': total_list}
stock_tmp = pd.DataFrame(stock_tmp)
stock_tmp = stock_tmp[:-1] ## Trim the last
stock_tmp = ccm.adjustDateRange(stock_tmp)
stock['new_tests'] = stock_tmp['new_tests'].values
return
def saveCsv_vaccinationByAge(self, mode='both'):
gr = ccm.GROUP_LATEST
if mode in ['data', 'both']:
stock = self.makeStock_vaccinationByAge()
data_l = {'key': ['latest_date'], 'value': [stock['latest_date']]}
data_l = pd.DataFrame(data_l)
data_r = self.addLabel_vaccinationByAge(stock)
data_r = pd.DataFrame(data_r)
## Save
name = '{}processed_data/{}/vaccination_by_age.csv'.format(ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_r)
name = '{}processed_data/{}/vaccination_by_age_label.csv'.format(ccm.DATA_PATH, gr)
ccm.saveCsv(name, data_l)
if mode in ['readme', 'both']:
self.makeReadme_vaccinationByAge(gr)
return