Ruby script for extracting the coordinates from the Colorado Flood Tweets. Generates file ~ 1mb. Includes integer dates for Tilemill styling.

extract_geo_tagged.rb

require 'json'
require 'csv'
require 'time'

file_path = '/path/to/2013ColoradoFloods.json'

#Open the CSV
CSV.open("/path/to/where/you/want/csv/geo_tagged_for_TM.csv", "w") do |csv|
	csv << ["Lon", "Lat", "Date", "SortDate", "User", "Text"]

	#Open the file and print the lines.
	cnt=0
	File.open( file_path ).each do |line|
	   tweet = JSON.load(line)
	   has_coords = tweet['coordinates'] or false
	   if has_coords
	   	coords=has_coords['coordinates']
	   	sort_time = Time.parse(tweet['created_at'])
	   	csv << coords + [tweet['created_at'], sort_time.tv_sec-1378000000, tweet['user']['screen_name'], tweet['text']]
	   end
	   if cnt%1000==0
	   	puts "#{cnt}: #{ coords }, #{tweet['user']['screen_name']}"
	   end
	   cnt+=1
	end
end