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 # Read infections table path and output type.
  21 import sys
  22 if len(sys.argv) != 3:
  23     print('Expecting infections table file path and output type as only arguments.')
  24     exit(1)
  25 infections_table = sys.argv[1]
  26 output_type = sys.argv[2]
  27
  28 # Read infections table file lines.
  29 f = open(infections_table, 'r')
  30 lines = f.readlines()
  31 f.close()
  32
  33 # Basic input validation.
  34 import datetime
  35 header_elements = lines[0].split()
  36 if set(header_elements) != district_pops.keys() or \
  37        len(header_elements) != len(district_pops.keys()):
  38     raise Exception('infections table: invalid header')
  39 line_count = 0
  40 for line in lines[1:]:
  41     line_count += 1
  42     fields = line.split()
  43     if len(header_elements) != len(fields) - 1:
  44         raise Exception('infections table: too many elements on line %s',
  45                         line_count)
  46     try:
  47         datetime.date.fromisoformat(fields[0])
  48     except ValueError:
  49         raise Exception('infections table: bad ISO date on line %s',
  50                         line_count)
  51     for field in fields[1:]:
  52         try:
  53             int(field)
  54         except ValueError:
  55             raise Exception('infections table: bad value on line %s',
  56                             line_count)
  57
  58 # Parse first table file line for the names and order of districts.
  59 db = {}
  60 sorted_districts = []
  61 for header in lines[0].split():
  62     sorted_districts += [header]
  63     db[header] = {}
  64
  65 # Seed DB with daily new infections data per district, per date.
  66 sorted_dates = []
  67 for line in lines[1:]:
  68     fields = line.split()
  69     date = fields[0]
  70     sorted_dates += [date]
  71     for i in range(len(sorted_districts)):
  72         district = sorted_districts[i]
  73         district_data = fields[i + 1]
  74         db[district][date] = {'new_infections': int(district_data)}
  75 sorted_dates.sort()
  76
  77 # In LaGeSo's data, the last "district" is actually the sum of all districts /
  78 # the whole of Berlin.
  79 #
  80 # Fail on any day where the "sum" district's new infections are not the proper
  81 # sum of the individual districts new infections.  Yes, sometimes Lageso sends
  82 # data that is troubled in this way.  It will then have to be fixed manually in
  83 # the table file, since we should have a human look at what mistake was
  84 # probably made.
  85 for date in sorted_dates:
  86     sum_district = sorted_districts[-1]
  87     day_sum = 0
  88     for district in sorted_districts[:-1]:
  89         day_sum += db[district][date]['new_infections']
  90     if day_sum != db[sum_district][date]['new_infections']:
  91         raise Exception('Questionable district infection sum in %s' % date)
  92
  93 # Enhance DB with data about weekly sums, averages, incidences per day.  Ignore
  94 # days that have less than 6 predecessors (we can only know a weekly average if
  95 # we have a whole week of data).
  96 for i in range(len(sorted_dates)):
  97     if i < 6:
  98         continue
  99     date = sorted_dates[i]
 100     week_dates = []
 101     for j in range(7):
 102         week_dates += [sorted_dates[i - j]]
 103     for district in sorted_districts:
 104         district_pop = district_pops[district]
 105         week_sum = 0
 106         for week_date in week_dates:
 107             week_sum += db[district][week_date]['new_infections']
 108         db[district][date]['week_sum'] = week_sum
 109         db[district][date]['week_average'] = week_sum / 7
 110         db[district][date]['week_incidence'] = (week_sum / district_pop) * 100000
 111
 112 # Optimized for web browser viewing.
 113 if output_type == 'html':
 114     print('<html>')
 115     print('<style>')
 116     print('table, tr, th, td { border: 1px solid black; }')
 117     print('</style>')
 118     print('<table>')
 119     print('<tr>')
 120     print('<th>date</th>')
 121     for district in sorted_districts:
 122         print('<th>%s</th>' % district)
 123     print('</tr>')
 124     sorted_dates.reverse()
 125     for date in sorted_dates:
 126         print('<tr>')
 127         print('<td>%s</td>' % date)
 128         for district in sorted_districts:
 129             district_data = db[district][date]
 130             week_sum = week_avg = week_inc = ''
 131             new_infections = district_data['new_infections']
 132             if 'week_sum' in district_data:
 133                 week_sum = '%s' % district_data['week_sum']
 134             if 'week_average' in district_data:
 135                 week_avg = '%.1f' % district_data['week_average']
 136             if 'week_incidence' in district_data:
 137                 week_inc = '%.1f' % district_data['week_incidence']
 138             print('<td>')
 139             print('<table>')
 140             print('<tr><th>new</th><td>%s</td></tr>' % new_infections)
 141             print('<tr><th>wsum</th><td>%s</td></tr>' % week_sum)
 142             print('<tr><th>wavg</th><td>%s</td></tr>' % week_avg)
 143             print('<tr><th>winc</th><td>%s</td></tr>' % week_inc)
 144             print('</table>')
 145             print('</td>')
 146         print('</tr>')
 147     print('</table>')
 148     print('</html>')
 149
 150 # Optimized for in-terminal curl.
 151 elif output_type == 'txt':
 152
 153     # Explain what this is.
 154     intro = \
 155 """Table of Berlin's Corona infection number development by districts.
 156 Updated daily around 9pm.
 157
 158 Abbrevations/explanations:
 159
 160 CW: Charlottenburg-Wilmersdorf
 161 FK: Friedrichshain-Kreuzberg
 162 Li: Lichtenberg
 163 MH: Marzahn-Hellersdorf
 164 Mi: Mitte
 165 Ne: Neukölln
 166 Pa: Pankow
 167 Re: Reinickendorf
 168 Sp: Spandau
 169 SZ: Steglitz-Zehlendorf
 170 TS: Tempelhof-Schöneberg
 171 TK: Treptow-Köpenick
 172 sum: sum for all the districts
 173 wsum: sum for last 7 days
 174 wavg: per-day average of new infections for last 7 days
 175 winc: incidence (x per 100k inhabitants) of new infections for last 7 days
 176
 177 Source code: https://plomlompom.com/repos/?p=berlin-corona-table
 178 """
 179     print(intro)
 180
 181     # Output table of enhanced daily infection data, newest on top,
 182     # separated into 7-day units.
 183     sorted_dates.reverse()
 184     weekday_count = 0
 185     sum_district = sorted_districts[-1]
 186     for date in sorted_dates:
 187
 188         # Week table header.
 189         if weekday_count == 0:
 190             print(' '*11, '  '.join(sorted_districts[:-1]),
 191                   sorted_districts[-1], 'wsum', ' wavg', 'winc')
 192             week_start_date = date
 193
 194         # Day data line.
 195         new_infections = []
 196         for district in sorted_districts:
 197             new_infections += [db[district][date]['new_infections']]
 198         week_sum = week_avg = week_inc = ''
 199         sum_district_data = db[sum_district][date]
 200         if 'week_sum' in sum_district_data:
 201             week_sum = '%4s' % sum_district_data['week_sum']
 202         if 'week_average' in sum_district_data:
 203             week_avg = '%5.1f' % sum_district_data['week_average']
 204         if 'week_incidence' in sum_district_data:
 205             week_inc = '%4.1f' % sum_district_data['week_incidence']
 206         print(date, ' '.join(['%3s' % infections
 207                               for infections in new_infections]),
 208               week_sum, week_avg, week_inc)
 209
 210         # Maintain 7-day cycle.
 211         weekday_count += 1
 212         if weekday_count != 7:
 213             continue
 214         weekday_count = 0
 215
 216         # After each 7 days, print summary for individual districts.
 217         weekly_sums = []
 218         weekly_avgs = []
 219         weekly_incs = []
 220         for district in sorted_districts[:-1]:
 221             weekly_sums += [db[district][week_start_date]['week_sum']]
 222             weekly_avgs += [db[district][week_start_date]['week_average']]
 223             weekly_incs += [db[district][week_start_date]['week_incidence']]
 224         print()
 225         print('district stats for week from %s to %s:' % (date, week_start_date))
 226         print(' '*7, '    '.join(sorted_districts[:-1]))
 227         print('wsum', ' '.join(['%5.1f' % wsum for wsum in weekly_sums]))
 228         print('wavg', ' '.join(['%5.1f' % wavg for wavg in weekly_avgs]))
 229         print('winc', ' '.join(['%5.1f' % winc for winc in weekly_incs]))
 230         print()