Save The Date (also, fun with historical weather data)

Laura and I would like to have the wedding outdoors. Naturally, this means that weather will play a significant factor. To get some idea of what the weather will be like on the big day, I whipped up some Python to calculate local average temperature, rainfall, and wind speed for the time surrounding April 5th.

import csv
import urllib2

STATION_ID = 'KNCOAKIS1'
WEATHER_URL = \
  'http://www.wunderground.com/weatherstation/WXDailyHistory.asp?'
REQUEST_DATA = {
  'ID': STATION_ID,
  'format': 1,
  'year': None,
  'month': None,
  'day': None,
  }

def _GetRequestData(year, month, day):
  request_data = REQUEST_DATA.copy()
  request_data['year'] = year
  request_data['month'] = month
  request_data['day'] = day
  return request_data

def GetDayStats(year, month, day):
  request_data = _GetRequestData(year, month, day)
  get_data = '&'.join('%s=%s' % (k, v) for k, v in request_data.items())
  url = WEATHER_URL + get_data
  data = urllib2.urlopen(url).read()
  # Have to clean up the data because Weather Underground doesn't return a
  # true CSV file. Instead, it has HTML markup in it :(
  data.strip()
  data = data.split('<br>')[:-1]
  # Remove trailing whitespace and comma.
  data = [r.strip()[:-1] for r in data]
  reader = csv.DictReader(data)
  averages = {
    'TemperatureF': 0,
    'WindSpeedMPH': 0,
    'WindSpeedGustMPH': 0,
    }
  sums = {
    'HourlyPrecipIn': 0,
    }
  mins = {
    'WindSpeedMPH': [],
    'WindSpeedGustMPH': [],
    }
  # DictReader has no __len__ defined. Have to keep count.
  num_rows = 0
  for row in reader:
    num_rows += 1
    for k in averages.keys():
    averages[k] += float(row[k])
    for k in sums.keys():
      sums[k] += float(row[k])
    for k in mins.keys():
      mins[k].append((row['Time'], float(row[k])))

  day_stats = {}
  for k, v in averages.iteritems():
    averages[k] = v / num_rows
  day_stats.update(averages)
  day_stats.update(sums)
  for k in mins.keys():
    values = mins[k][:]
    values.sort(key=lambda x: x.__getitem__(1))
    day_stats['Min%s' % k] = values[0]
  return day_stats

def GetDayHistory(month, day, start_year, end_year):
  days = []
  for year in range(start_year, end_year + 1):
    days.append(GetDayStats(year, month, day))
  avg_stats = {}
  min_times = []
  for stats in days:
    for k, v in stats.iteritems():
      if isinstance(v, tuple):
        min_times.append(v[0])
        v = v[1]
      avg_stats[k] = avg_stats.get(k, 0) + v
  for k, v in avg_stats.iteritems():
    avg_stats[k] = v / len(days)
  return avg_stats, min_times

if __name__ == '__main__':
  month = 4
  for day in range(1, 8):
    print 'History for %d/%d.' % (month, day)
    print GetDayHistory(month, day, 2003, 2007)

Here's the results for April 5th over the past 5 years:

History for 4/5.
({
  'TemperatureF': 61.136611720960197,
  'WindSpeedMPH': 7.3304707160832763,
  'WindSpeedGustMPH': 11.706614130829021,
  'MinWindSpeedGustMPH': 0.0,
  'MinWindSpeedMPH': 0.0,
  'HourlyPrecipIn': 0.0
}, 
['2003-04-05 00:30:02', '2003-04-05 01:30:03',
'2004-04-05 04:45:01', '2004-04-05 20:15:01',
'2005-04-05 02:30:04', '2005-04-05 03:00:01',
'2006-04-05 20:45:01', '2006-04-05 14:45:01',
'2007-04-05 21:30:00', '2007-04-05 22:45:00'])

From this data, it's obvious that rain and temperature shouldn't be a problem. The biggest problem will be the wind. Here are the relevant weather charts for April 5th, 2007 (unfortunately, 2008 is not available).



Clearly the best times are in the morning and the evening. The evening is probably best since it's about 10 degrees warmer at 20:00 than 8:00 and the wind speed drops steadily after 19:00.