plomlompom.com Git - berlin-corona-table/blob - enhance_table.py

   1 #!//usr/bin/env python3
   2
   3 # District population numbers as per Wikipedia.
   4 district_pops = {
   5   'CW': 342332,
   6   'FK': 289762,
   7   'Li': 291452,
   8   'MH': 268548,
   9   'Mi': 384172,
  10   'Ne': 329691,
  11   'Pa': 407765,
  12   'Re': 265225,
  13   'Sp': 243977,
  14   'SZ': 308697,
  15   'TS': 351644,
  16   'TK': 271153,
  17   'sum': 3754418,
  18 }
  19
  20 f = open('daily_infections_table.txt', 'r')
  21 lines = f.readlines()
  22 f.close()
  23
  24 # Parse first table file line for the names and order of districts.
  25 db = {}
  26 sorted_districts = []
  27 for header in lines[0].split():
  28     sorted_districts += [header]
  29     db[header] = {}
  30
  31 # Seed DB with daily new infections data per district, per date.
  32 sorted_dates = []
  33 for line in lines[1:]:
  34     fields = line.split()
  35     date = fields[0]
  36     sorted_dates += [date]
  37     for i in range(len(sorted_districts)):
  38         district = sorted_districts[i]
  39         district_data = fields[i + 1]
  40         db[district][date] = {'new_infections': int(district_data)}
  41 sorted_dates.sort()
  42
  43 # Fail on any day where the "sum" district's new infections are not the proper
  44 # sum of the individual districts new infections.  Yes, sometimes Lageso sends
  45 # data that is troubled in this way.  It will then have to be fixed manually in
  46 # the table file, since we should have a human look at what mistake was
  47 # probably made.
  48 for date in sorted_dates:
  49     sum_district = sorted_districts[-1]
  50     day_sum = 0
  51     for district in sorted_districts[:-1]:
  52         day_sum += db[district][date]['new_infections']
  53     if day_sum != db[sum_district][date]['new_infections']:
  54         raise Exception('Questionable district infection sum in %s' % date)
  55
  56 # Enhance DB with data about weekly sums, averages, incidences per day.  Ignore
  57 # days that have less than 6 predecessors (we can only know a weekly average if
  58 # we have a whole week of data).
  59 for i in range(len(sorted_dates)):
  60     if i < 6:
  61         continue
  62     date = sorted_dates[i]
  63     week_dates = []
  64     for j in range(7):
  65         week_dates += [sorted_dates[i - j]]
  66     for district in sorted_districts:
  67         district_pop = district_pops[district]
  68         week_sum = 0
  69         for week_date in week_dates:
  70             week_sum += db[district][week_date]['new_infections']
  71         db[district][date]['week_sum'] = week_sum
  72         db[district][date]['week_average'] = week_sum / 7
  73         db[district][date]['week_incidence'] = (week_sum / district_pop) * 100000
  74
  75 # Explain what this is.
  76 intro = """Table of Berlin's Corona infection number development by districts.
  77 Updated daily around 9pm.
  78
  79 Abbrevations/explanations:
  80
  81 CW: Charlottenburg-Wilmersdorf
  82 FK: Friedrichshain-Kreuzberg
  83 Li: Lichtenberg
  84 MH: Marzahn-Hellersdorf
  85 Mi: Mitte
  86 Ne: Neukölln
  87 Pa: Pankow
  88 Re: Reinickendorf
  89 Sp: Spandau
  90 SZ: Steglitz-Zehlendorf
  91 TS: Tempelhof-Schöneberg
  92 TK: Treptow-Köpenick
  93 sum: sum for all the districts
  94 wsum: sum for last 7 days
  95 wavg: per-day average of new infections for last 7 days
  96 winc: incidence (x per 100k inhabitants) of new infections for last 7 days
  97
  98 Source code: https://plomlompom.com/repos/?p=berlin-corona-table
  99 """
 100 print(intro)
 101
 102 # Output table of enhanced daily infection data, newest on top, separated into
 103 # 7-day units.
 104 sorted_dates.reverse()
 105 weekday_count = 0
 106 for date in sorted_dates:
 107
 108     # Week table header.
 109     if weekday_count == 0:
 110         print(' '*11, '  '.join(sorted_districts[:-1]),
 111               sorted_districts[-1], 'wsum', ' wavg', 'winc')
 112         week_start_date = date
 113
 114     # Day data line.
 115     new_infections = []
 116     for district in sorted_districts:
 117         new_infections += [db[district][date]['new_infections']]
 118     week_sum = week_avg = week_inc = ''
 119     sum_district = sorted_districts[-1]
 120     sum_district_data = db[sum_district][date]
 121     if 'week_sum' in sum_district_data:
 122         week_sum = '%4s' % sum_district_data['week_sum']
 123     if 'week_average' in sum_district_data:
 124         week_avg = '%5.1f' % sum_district_data['week_average']
 125     if 'week_incidence' in sum_district_data:
 126         week_inc = '%4.1f' % sum_district_data['week_incidence']
 127     print(date, ' '.join(['%3s' % infections for infections in new_infections]),
 128           week_sum, week_avg, week_inc)
 129
 130     # Maintain 7-day cycle.
 131     weekday_count += 1
 132     if weekday_count != 7:
 133         continue
 134     weekday_count = 0
 135
 136     # After each 7 days, print summary for individual districts.
 137     weekly_sums = []
 138     weekly_avgs = []
 139     weekly_incs = []
 140     for district in sorted_districts[:-1]:
 141         weekly_sums += [db[district][week_start_date]['week_sum']]
 142         weekly_avgs += [db[district][week_start_date]['week_average']]
 143         weekly_incs += [db[district][week_start_date]['week_incidence']]
 144     print()
 145     print('district stats for week from %s to %s:' % (date, week_start_date))
 146     print(' '*7, '    '.join(sorted_districts[:-1]))
 147     print('wsum', ' '.join(['%5.1f' % wsum for wsum in weekly_sums]))
 148     print('wavg', ' '.join(['%5.1f' % wavg for wavg in weekly_avgs]))
 149     print('winc', ' '.join(['%5.1f' % winc for winc in weekly_incs]))
 150     print()