plotti

def total_edge_weight(D):
        total = 0
        for edge in D.edges(data=True):
                total += edge[2]["weight"]
        return total

def average_tie_strength(D):
        return float(total_edge_weight(D))/len(D.edges())

D = nx.DiGraph()

plotti

def reciprocity(D):
        G=D.to_undirected() # copy
        for (u,v) in D.edges():
            if not D.has_edge(v,u):
                    G.remove_edge(u,v)
        return float(len(G.edges()))/len(D.to_undirected().edges())

# A simple test:

G = nx.DiGraph()

plotti

  def self.collect_list_memberships(username, project_id = "")
    page_size = LIST_PAGE_SIZE
    attempts = 0
    while Project.get_remaining_hits == "timeout"
      puts "collect_list_memberships waiting..."
      sleep(60) 
    end
    begin
      result = @@twitter.memberships(username, {:cursor => -1, :count => page_size})
      lists = result["lists"]

plotti

  def self.collect_list_members(username, list_id,project_id)    
    while Project.get_remaining_hits == "timeout"
      puts "collect_list_members waiting..."
      sleep(60) 
    end
    result = @@twitter.list_members(username, list_id, {:cursor => -1})    
    members = result["users"]
    next_cursor = result["next_cursor"]
    old_next_cursor = 0    
    

plotti

  def generate_most_listed_members
      seen_lists = []
      seen_membersets = []
      @tmp_persons = []
      self.persons.each do |person|
        #puts person.username
        @tmp_persons << {:username => person.username, :list_count => 1,
          :uri => "http://www.twitter.com/#{person.username}", :followers => person.followers_count,
          :friends => person.friends_count}
      end

plotti

Username,Followers,List Count,URI
aplusk,10137792,343,http://www.twitter.com/aplusk
JimCarrey,7159113,297,http://www.twitter.com/JimCarrey
tomhanks,3773553,288,http://www.twitter.com/tomhanks
ActuallyNPH,3061975,283,http://www.twitter.com/ActuallyNPH
NathanFillion,1222189,232,http://www.twitter.com/NathanFillion
RedHourBen,2655373,230,http://www.twitter.com/RedHourBen
peterfacinelli,1937513,215,http://www.twitter.com/peterfacinelli
Alyssa_Milano,2016472,214,http://www.twitter.com/Alyssa_Milano
KevinSpacey,2259294,210,http://www.twitter.com/KevinSpacey

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

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

airlines_aviation
army_military_veteran
astronomy_physics
beauty_fashion_shopping
career_employment
charity_philanthropy
etsy_handmade
exercise_fitness
finance_economics
healthcare_medicine

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