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