Travis County precinct-by-precinct results parser

travis-elex-formatter.py

import csv, sys

# This is a quick and dirty solution to walk the file that Travis County distributes
# the day after the election and create a new .csv file that has one row per race
# per precinct with the winner and his/her/its vote percentage
#
# CLI usage: python travis-elex-formatter.py input_file.csv output_file.csv

last_field = None
winner_total_votes = 0
running_vote_total = 0
race_results = []

with open(sys.argv[1], 'rb') as input_file, open(sys.argv[2], 'wb') as output_file:
  # Open a reader for the input file
  results = csv.reader(input_file, delimiter=',')
  # Skip row headers for the input row
  next(results, None)
  # And open a write for the output file
  output = csv.writer(output_file)
  # Write row headers
  output.writerow(['Precinct', 'Race', 'Candidate name', 'Vote count', 'Vote share', 'Won', 'Race tooltip'])
  # Loop over the input, parse & write the new file
  for row in results:
    precinct = row[0]
    candidate_name = row[14]
    total_votes = int(row[19])
    race = row[10]

    # Generate a field name that is the precinct + the race name; When this
    # changes, we know we're on a new race
    field = precinct + race

    # If we're now in a different race, different precinct then save the winner
    # before moving on
    if (field != last_field) and (winner_total_votes != 0):
      winner_percent_share = float(winner_total_votes) / float(running_vote_total)
      won_race = winner_percent_share > .5

      # Build the tooltip text
      sorted_races = []
      for sorted_race in sorted(race_results, key=lambda k: k['votes'], reverse=True):
        percent = float(sorted_race['votes']) / float(running_vote_total) * 100

        sorted_races.append("%s: %s (%.1f%%)" % (sorted_race['candidate'], sorted_race['votes'], percent))

      # Write a new row to the results file
      output.writerow([
        winner_precinct,
        winner_race,
        winner_candidate_name,
        winner_total_votes,
        winner_percent_share,
        won_race,
        "\n".join(sorted_races)
      ])

      # Reset variables
      winner_total_votes = 0
      running_vote_total = 0
      race_results = []

    # If this is the first for this precinct or the current row has a candidate
    # with a higher total than the last, mark a new winner
    if total_votes > winner_total_votes:
      winner_candidate_name = candidate_name
      winner_total_votes = total_votes
      winner_precinct = precinct
      winner_race = race

    # Add this candidate to the list of results for this race in this precinct
    race_results.append({
      'candidate': candidate_name,
      'votes': total_votes
    })

    # Keep a running vote total to calculate the percentage down the road
    running_vote_total = running_vote_total + total_votes

    # Save this field name for the next row so we can tell then if we're starting
    # a new one
    last_field = field

print "Done"