def activity_stat(path, usernames, ids, adj):
rt_c = np.zeros((len(ids), len(ids)))
qt_c = np.zeros((len(ids), len(ids)))
rep_c = np.zeros((len(ids), len(ids)))
node_w = np.zeros((len(ids)))
for name in usernames:
tweets = functions.load_json_list(path+name+'.json')
for item in tweets:
if 'retweeted_status' in item.keys() and item['retweeted_status']['user']['id_str'] in ids:
rt_c[usernames.index(name)][ids.index(item['retweeted_status']['user']['id_str'])]+=1
#if its a quoted tweet from and account in the friend net then update qt_count value
if item['is_quote_status']==True and 'quoted_status' in item.keys():
if item['quoted_status']['user']['id_str'] in ids:
qt_c[usernames.index(name)][ids.index(item['quoted_status']['user']['id_str'])]+=1
#if its a reply(mention) tweet to an account from the friend net then update rep_count value
if item['in_reply_to_user_id'] != None and str(item['in_reply_to_user_id']) in ids:
rep_c[usernames.index(name)][ids.index(str(item['in_reply_to_user_id']))]+=1
if len(tweets)>0:
item1=tweets[-1]
item2=tweets[0]
idx = usernames.index(name)
node_w[idx] = nodes_stat(item1, item2, len(tweets))
else:
node_w[idx] = 0
np.fill_diagonal(rt_c, 0)
np.fill_diagonal(qt_c, 0)
np.fill_diagonal(rep_c, 0)
return rt_c, qt_c, rep_c, node_w
def read_multiple_net(accounts, adr):
data, ids, usernames, color = [], [], [], []
counter = 0
for acc in accounts:
path = adr+'/'+acc
#read data for current user
raw_data = functions.load_json_list(path+'/'+acc+'_complete.txt')
base_usernames = functions.load(path+'/'+acc+'_usernames.txt')
base_ids = functions.load(path+'/'+acc+'_ids.txt')
#substitution usernames with user ids for current user
temp_data, temp_ids, temp_usernames = prepare_data(raw_data, base_ids, base_usernames)
data, ids, usernames, color = update_data(data,
ids,
usernames,
temp_data,
temp_ids,
temp_usernames,
color)
counter+=1
color = update_color(color, len(data), counter)
return data, ids, usernames, color
username = '******'
save = False
input_dir = 'TWEETS DIRECTORY'
output_dir = 'RESULT DIRECTORY'
directory = os.fsencode(input_dir)
num_of_files = len([name for name in os.listdir(directory)])
counter = 0
english_counter = 0
tweet_counter = 0
for file in os.listdir(directory):
file_name = os.fsdecode(file)
path = input_dir + file_name
data = functions.load_json_list(path)
tweet_counter += len(data)
counter += 1
for item in data:
if item['lang'] == 'en':
english_counter += 1
sys.stdout.write('\r%d/%d' % (counter, num_of_files))
print("\n=========", username, "=========")
print("\nnumber of tweets;", tweet_counter)
print("found hashtags: ", english_counter)
print("final value:", round(english_counter / tweet_counter, 5))
user = '******'
alfa = 0.4
extracting_weights = False
plot = True
disjoint = False
bc_analysis = False
cc_analysis = False
jsc = False
sdi = False
entropy = False
w = [0.2, 0.5, 0.3] #ret, qt, rep
path = 'EGOS DIRECTORY' + user
raw_data = functions.load_json_list(path + '/' + user + '_complete.txt')
base_usernames = functions.load(path + '/' + user + '_usernames.txt')
base_ids = functions.load(path + '/' + user + '_ids.txt')
data, ids, usernames = info_diff_func.prepare_data(raw_data, base_ids,
base_usernames)
print("\n===========", user, "===========\n")
#creating adjacency matrix
adj = info_diff_func.adjacency(ids, data)
#a simple directed graph using networkx
G = info_diff_func.simple_graph(adj)
#extracting links