plotti

Person 1 archaeologymag. Time per person: 7.783559. 
Person 2 ArchaeologyDN. Time per person: 1.097274. 
Person 3 archaeologynews. Time per person: 0.085687.
Person 4 NEAarchaeology. Time per person: 0.184254. 
Person 5 CurrentArchaeo. Time per person: 0.116461.

plotti

  def find_solr_at_connections
    users = {}    
    persons.each do |person|        
	users[person.id] = person.username
    end
    values = []
    i = 0
    persons.each do |person|
      i += 1
      t1 = Time.now

plotti

  # Indexing the text field with solr
  searchable do
    text :text
    integer :person_id
    time :published_at
    text :retweeters do 
    	retweet_ids.collect{|retweet| retweet[:id]}
    end
  end

plotti

# Helper method that is used when creating delayed jobs.
  def wait_for_jobs(jobname)
  continue = true
    while continue
      found_pending_jobs = 0
      Delayed::Job.all.each do |job|
        if job.handler.include? jobname
              found_pending_jobs += 1
        end
        if job.attempts >= 4

plotti

 # A version that used delayed jobs to split the work among a lot of workers
  # Deprecated because the bottleneck was querying the DB
  # Same as find all valued connections only trying to make it faster
  def find_delayed_at_connections(friend = true, follower = false, category = false)            
    usernames = persons.collect{|p| p.username}
    persons.each do |person|
      Delayed::Job.enqueue(AggregateAtConnectionsJob.new(person.id,self.id,usernames))
    end    
    wait_for_jobs("AggregateAtConnectionsJob")
    self.return_delayed_at_connections

plotti

  def find_at_connections()
    values = []
    usernames = persons.collect{|p| p.username}
    i = 0
    persons.each do |person|
      t1 = Time.now      
      i += 1      
      person.feed_entries.each do |tweet|        
        #puts "#Analyzing tweet #{tweet.id}"
        usernames.each do |tmp_user|

plotti

#Second step is to output the final partitions according to the ranking of the persons in their groups
seen_persons = []
final_candidates = {}
seen_projects = []
@@communities.each do |community|
  project = Project.find(community)      
  if merged[project.name] == nil
    project_members = members[project.name]
    project_name = project.name
  else    

plotti

airlines_aviation
army_military_veteran
astronomy_physics
beauty_fashion_shopping
career_employment
charity_philanthropy
etsy_handmade
exercise_fitness
finance_economics
healthcare_medicine

plotti

#Define how many list places should be considered
MAX = 200

#Threshold: The threshold until which the categories should be merged (e.g. 0.2 = 20 % of members are shared)
THRESHOLD = 0.2

outfile = CSV.open("data/partitions#{MAX}_#{THRESHOLD}.csv", "wb")
final_partition = CSV.open("data/final_partitions#{MAX}_#{THRESHOLD}.csv", "wb")
outfile << ["Name","Original Category", "Original Category Place", "Assigned Category", "Assigned Category Place", "Competing Categories", "Details"]

plotti

Username,Followers,List Count,URI
aplusk,9777167,83,http://www.twitter.com/aplusk
katyperry,16107303,80,http://www.twitter.com/katyperry
ladygaga,20799180,79,http://www.twitter.com/ladygaga
TheEllenShow,9922576,75,http://www.twitter.com/TheEllenShow
RyanSeacrest,6176522,73,http://www.twitter.com/RyanSeacrest
tomhanks,3587554,73,http://www.twitter.com/tomhanks
ActuallyNPH,2908961,68,http://www.twitter.com/ActuallyNPH
ConanOBrien,5213126,63,http://www.twitter.com/ConanOBrien
Oprah,10431484,62,http://www.twitter.com/Oprah