def bot_participation():
Bot = bot.load_bot()
dir_name = "RetweetNew/"
files = os.listdir(dir_name)
bot_list = []
for postid in files:
with open(dir_name + postid, 'r') as f:
tweets = json.load(f)
users = [tweet['user'] for tweet in tweets.values()]
bots = [bot.check_bot(Bot, user) for user in users]
bot_list.append(bots.count(1) / bots.count(0))
box = BoxPlot(1)
box.set_data(bot_list, '')
box.set_xticks('bot_ratio')
box.save_image('Image/bot_ratio_box.png')
def time_to_depth_echo_chamber(filename):
_, _, time_depth, _, user_depth = get_depth_time_series(None)
print(len(time_depth))
#with open('Data/time_series_data.json', 'w') as f:
# json.dump({'time_depth' : time_depth, 'user_depth' : user_depth}, f)
#with open('Data/time_series_data.json', 'r') as f:
# data = json.load(f)
#time_depth = data['time_depth']
#user_depth = data['user_depth']
print("time series data load done ")
echo_chamber_values = {}
non_echo_chamber_values = {}
for item in ['time_depth', 'user_depth']:
echo_chamber_values[item] = {}
non_echo_chamber_values[item] = {}
for i in range(1,20):
echo_chamber_values[item][i] = []
non_echo_chamber_values[item][i] = []
Bot = bot.load_bot()
echo_chamber_cascade_root = {}
cascade_veracity = {}
echo_chamber_users = e_util.get_echo_chamber_users(filename)
files = os.listdir('RetweetNew')
#collect echo chamber user participate cascade
#for postid in echo_chamber_users.keys():
for postid in files:
v = veracity_type(postid).title()
#get origin tweet of echo chamber user
with open('RetweetNew/%s'%postid, 'r') as f:
tweets = json.load(f)
for tweet in tweets.values():
try:
#if tweet['user'] in echo_chamber_users[postid].keys():
origin = tweet['origin']
otid = tweet['origin_tweet']
#if origin in echo_chamber_users[postid].keys():
if tweet['user'] in echo_chamber_users[postid].keys():
echo_chamber_cascade_root[tweet['origin_tweet']] = 1
except KeyError :
pass
cascade_veracity[tweet['origin_tweet']] = v
print("echo chamber cascade extraction done")
echo_chamber_cascades = echo_chamber_cascade_root.keys()
print('echo chamber cascades')
#print(echo_chamber_cascades)
e = {}; n = {}; r = {}; #echo, non echo, ranked echo
for item in ['True', 'False', 'Mixed']:
e[item] = {}
n[item] = {}
r[item] = {}
for d_type in ['user_depth', 'time_depth']:
e[item][d_type] = {}
n[item][d_type] = {}
r[item][d_type] = {}
for i in range(1, 20):
e[item][d_type][i] = []
n[item][d_type][i] = []
r[item][d_type][i] = []
for key in time_depth.keys():
v = cascade_veracity[key]
if v !='True' and v != 'False':
v = 'Mixed'
if key in echo_chamber_cascades:
#for i in range(1, max(time_depth[key].keys())+1):
for i in range(1, max(time_depth[key].keys())+1):
try:
echo_chamber_values['time_depth'][i].append(time_depth[key][i])
echo_chamber_values['user_depth'][i].append(user_depth[key][i])
e[v]['time_depth'][i].append(time_depth[key][i])
e[v]['user_depth'][i].append(user_depth[key][i])
except KeyError:
pass
else:
for i in range(1, max(time_depth[key].keys())+1):
try :
non_echo_chamber_values['time_depth'][i].append(time_depth[key][i])
non_echo_chamber_values['user_depth'][i].append(user_depth[key][i])
n[v]['time_depth'][i].append(time_depth[key][i])
n[v]['user_depth'][i].append(user_depth[key][i])
except KeyError:
pass
box = BoxPlot(1)
box.set_multiple_data([echo_chamber_values['time_depth'], non_echo_chamber_values['time_depth']])
box.set_ylog()
box.set_label('Depth', 'Minutes to Depth')
box.save_image('%s/time_depth_echo_chamber_box.png'%foldername)
print(echo_chamber_values['time_depth'])
#draw time to depth, user to depth of cascade for echo chamber users participated or non echo chamer users participated
with open('Data/Figure/5_2_1.json', 'w') as f:
json.dump([echo_chamber_values['time_depth'], non_echo_chamber_values['time_depth']], f)
draw_time_to_depth_echo_chamber([echo_chamber_values['time_depth'], non_echo_chamber_values['time_depth']], ['echo chamber', 'no echo chamber'], 'median minutes', 'time_depth_echo_chamber_line')
draw_time_to_depth_echo_chamber([echo_chamber_values['user_depth'], non_echo_chamber_values['user_depth']], ['echo chamber', 'no echo chamber'], 'median unique users', 'user_depth_echo_chamber_line')
with open('Data/Figure/5_2_time.json', 'w') as f:
json.dump({'e':echo_chamber_values['time_depth'][1], 'ne':non_echo_chamber_values['time_depth'][1]}, f)
#draw cdf with top retweet
cdf = CDFPlot()
cdf.set_label('Propagation Time', 'CDF')
cdf.set_log(True)
#cdf.set_ylog()
cdf.set_data(echo_chamber_values['time_depth'][1], '')
cdf.set_data(non_echo_chamber_values['time_depth'][1], '')
cdf.save_image('Image/20181105/depth_propagation_time_cdf.png')
"""
def cascade_depth_distribution():
c_breadth, c_depth, c_unique_users = e_util.get_cascade_max_breadth()
depth_cascade = {}
depth_unique_users = {}
depth_breadth = {}
cascade_list = {}
for i in range(1,20):
depth_cascade[i] = []
depth_unique_users[i] = []
depth_breadth[i] = []
for ccc, postid in enumerate(rumors):
#if postid != '126119':
# continue
#print(postid)
with open(dirname + '/' + postid, 'r') as f:
tweets = json.load(f)
for tweet in tweets.values():
otid = tweet['origin_tweet']
if cascade_list.get(otid, None) == None:
max_depth = c_depth[otid]
users = c_unique_users[otid]
max_breadth = c_breadth[otid]
# print(otid, max_depth)
cascade_list[otid] = 1
depth_cascade[max_depth].append(tweet['cascade'])
depth_unique_users[max_depth].append(users)
depth_breadth[max_depth].append(max_breadth)
# if ccc > 10:
# break
print(ccc)
#print(depth_cascade)
cascade_list = []
user_list = []
breadth_list = []
#box.set_multiple_data([e_child, ne_child])
for i in range(1, 18):
cascade_list.append(depth_cascade[i])
user_list.append(depth_unique_users[i])
breadth_list.append(depth_breadth[i])
box = BoxPlot(1)
box.set_data(cascade_list, '')
box.set_ylog()
box.set_label('Depth', 'Cascade Size')
box.save_image('Image/%s/depth_cascade_dist.png'%foldername)
box = BoxPlot(1)
box.set_data(user_list, '')
box.set_ylog()
box.set_label('Depth', 'Number of Users')
box.save_image('Image/%s/depth_user_dist.png'%foldername)
box = BoxPlot(1)
box.set_data(breadth_list, '')
box.set_ylog()
box.set_label('Depth', 'Breadth Size')
box.save_image('Image/%s/depth_breadth_dist.png'%foldername)
def echo_chamber_diversity(filename):
Bot = bot.load_bot()
dirname = 'Retweet/'
files = os.listdir(dirname)
if filename == None:
echo_chamber_users = {}
for postid in files:
echo_chamber_users[postid] = {}
else:
echo_chamber_users = e_util.get_echo_chamber_users(filename)
echo_tweet_diversity = []; echo_source_diversity = [];
necho_tweet_diversity = []; necho_source_diversity = [];
for postid in files:
with open(dirname + postid) as f:
tweets = json.load(f)
non_echo_users = {}
for tweet in tweets.values():
user = tweet['user']
#non echo chamber collect
if not user in echo_chamber_users[postid]:
non_echo_users[user] = 1
print(len(echo_chamber_users[postid]), len(non_echo_users))
timeline_dir = '../Timeline/'
#collect echo chamber users' source diversity
err = 0; nerr = 0
for user in echo_chamber_users[postid]:
try:
with open(timeline_dir + user, 'r') as f:
user_tweets = json.load(f)
except IOError as e:
#print(e)
err +=1
continue
tweet_diversity, source_diversity = get_diversity(user_tweets)
if tweet_diversity != None:
echo_tweet_diversity.append(tweet_diversity)
if source_diversity != None:
echo_source_diversity.append(source_diversity)
for user in non_echo_users:
try:
with open(timeline_dir + user, 'r') as f:
user_tweets = json.load(f)
except IOError as e:
#print(e)
nerr += 1
continue
tweet_diversity, source_diversity = get_diversity(user_tweets)
if tweet_diversity != None:
necho_tweet_diversity.append(tweet_diversity)
if source_diversity != None:
necho_source_diversity.append(source_diversity)
#print(err, nerr)
#break
#CDF
cdf = CDFPlot()
cdf.set_label('Retweet Origin Diversity', 'CDF')
#cdf.set_log(True)
cdf.set_data(echo_tweet_diversity, 'Echo Chamber')
cdf.set_data(necho_tweet_diversity, 'Non Echo Chamber')
cdf.set_legends(['Echo CHamber', 'Non Echo CHamber'], 'User Type')
cdf.save_image('Image/20181002/source_diversity_retweet_cdf.png')
cdf = CDFPlot()
cdf.set_label('Source News Diversity', 'CDF')
#cdf.set_log(True)
cdf.set_data(echo_source_diversity, 'Echo Chamber')
cdf.set_data(necho_source_diversity, 'Non Echo Chamber')
cdf.set_legends(['Echo CHamber', 'Non Echo CHamber'], 'User Type')
cdf.save_image('Image/20181002/source_diversity_news_cdf.png')
#BoxPlot
box = BoxPlot(1)
box.set_data([echo_tweet_diversity, necho_tweet_diversity],'')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber', 'All'])
box.set_label('', 'Retweet Origin Diversity')
box.save_image('Image/20181002/source_diversity_retweet.png')
box = BoxPlot(1)
box.set_data([echo_source_diversity, necho_source_diversity],'')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber', 'All'])
box.set_label('', 'Source News Diversity')
box.save_image('Image/20181002/source_diversity_news.png')
"'2016-01-01' and '2016-12-31'", "'2017-01-01' and '2017-12-31'"
]
for date in year:
sharecount_cdf("between " + date, date[1:5])
category_list = category_list()
all_sharecount_list = []
for item in category_list:
#sharecount list of one category
all_sharecount_list.append(sharecount_by_category(item))
subplot_num = lambda x: int(math.sqrt(x)) if math.sqrt(x).is_integer(
) else int(math.sqrt(x)) + 1
#Display BoxPlot and save at the path
BoxPlt = BoxPlot(subplot_num(25))
for i in range(25):
BoxPlt.set_data(all_sharecount_list[i], '')
BoxPlt.set_title(category_list[i])
BoxPlt.set_ylim(1000)
BoxPlt.save_image('./image/sharecount_box_plot.png')
all_sharecount_list = []
#category with source presence
for item in category_list:
all_sharecount_list.append([
sharecount_by_category_with_source(item),
sharecount_by_category_without_source(item)
])
def draw_propagation_time_to_group():
print('echo chamber 2')
echo_v2, necho_v2, recho_v2, rnecho_v2 = propagation_time_to_group(
'Data/echo_chamber2.json')
box = BoxPlot(1)
box.set_data([echo_v2, necho_v2], '')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group2.png' % folder)
box = BoxPlot(1)
box.set_data([recho_v2, rnecho_v2], '')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group_r2.png' % folder)
print('echo chamber 3')
echo_v3, necho_v3, recho_v3, rnecho_v3 = propagation_time_to_group(
'Data/echo_chamber3.json')
box = BoxPlot(1)
box.set_data([echo_v3, necho_v3], '')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group3.png' % folder)
box = BoxPlot(1)
box.set_data([recho_v3, rnecho_v3], '')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group_r3.png' % folder)
print('echo chamber 4')
echo_v4, necho_v4, recho_v4, rnecho_v4 = propagation_time_to_group(
'Data/echo_chamber4.json')
box = BoxPlot(1)
box.set_data([echo_v4, necho_v4], '')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group4.png' % folder)
box = BoxPlot(1)
box.set_data([recho_v4, rnecho_v4], '')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group_r4.png' % folder)
_, non_echo, _, rnon_echo = propagation_time_to_group(None)
print(len(echo_v2), len(echo_v3), len(echo_v4), len(non_echo))
box = BoxPlot(1)
box.set_data([echo_v2, echo_v3, echo_v4, non_echo], '')
box.set_xticks(['Echo Chamber2', 'Echo Chamber3', 'Echo Chamber4', 'All'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group.png' % folder)
box = BoxPlot(1)
box.set_data([recho_v2, recho_v3, recho_v4, rnon_echo], '')
box.set_xticks(['Echo Chamber2', 'Echo Chamber3', 'Echo Chamber4', 'All'])
box.set_label('', 'Propagation Time')
box.save_image('Image/%s/propagation_time_to_group_r.png' % folder)
def draw_propagation_velocity():
echo_v2, _, echo_p2, necho_p2 = rumor_propagation_velocity(
'Data/echo_chamber2.json')
#echo_v3, _ = rumor_propagation_velocity('Data/echo_chamber3.json')
#echo_v4, _ = rumor_propagation_velocity('Data/echo_chamber4.json')
_, non_echo, _, _ = rumor_propagation_velocity(None)
#print(len(echo_v2), len(echo_v3), len(echo_v4), len(non_echo))
box = BoxPlot(1)
box.set_data([echo_v2, non_echo], '')
box.set_xticks(['Echo Chamber2', 'All'])
#box.set_data([echo_v2, echo_v3, echo_v4, non_echo],'')
#box.set_xticks(['Echo Chamber2', 'Echo Chamber3', 'Echo Chamber4', 'All'])
box.set_label('', 'Mean Propagation Time')
box.save_image('Image/%s/propagation_time.png' % folder)
box = BoxPlot(1)
box.set_multiple_data([echo_p2, necho_p2])
box.set_ylog()
box.set_label('Depth', 'Propagation Time')
box.save_image('Image/%s/child_all_time_propagation.png' % folder)
def propagation_parent_to_child():
Bot = bot.load_bot()
dirname = 'RetweetNew/'
files = os.listdir(dirname)
filename = 'Data/echo_chamber2.json'
if filename == None:
echo_chamber_users = {}
for postid in files:
echo_chamber_users[postid] = {}
else:
echo_chamber_users = e_util.get_echo_chamber_users(filename)
echo_chamber_cascades = {}
tweet_cache = {}
'''
for postid in echo_chamber_users.keys():
users = echo_chamber_users[postid] #echo chamber users
with open('RetweetNew/' + postid, 'r') as f:
tweets = json.load(f)
tweet_cache[postid] = tweets
for tweet in tweets.values():
if tweet['user'] in users:
root_id = tweet['origin_tweet'] #root tweet id
echo_chamber_cascades[root_id] = 1
echo_chamber_cascades_ids = echo_chamber_cascades.keys()
'''
#print(echo_chamber_cascades_ids)
e_child = {}
ne_child = {}
e_time = {}
ne_time = {}
ne_time2 = {}
for i in range(1, 20):
e_child[i] = []
ne_child[i] = []
e_time[i] = {}
ne_time[i] = {}
ne_time2[i] = {}
print(len(echo_chamber_users.keys()))
for ccc, postid in enumerate(files):
#if postid != '150232' and postid != '29947':
# continue
with open(dirname + postid, 'r') as f:
tweets = json.load(f)
#tweets = tweet_cache[postid]
#if not util.is_politics(postid):
#if not util.is_non_politics(postid):
#if not util.is_veracity(postid, 'False'):
#if not util.is_veracity(postid, 'Mixture,Mostly False,Mostly True'):
# continue
#order by timeline
sort = {}
for key in tweets.keys():
tweet = tweets[key]
sort[key] = parser.parse(tweet['time'])
#sort by time
new_list = sorted(sort.items(), key=lambda x: x[1])
sorted_ids = [item[0] for item in new_list]
e_users = echo_chamber_users[postid]
#e_users = echo_chamber_users.get(postid, [])
print(len(e_users))
for i, tid in enumerate(sorted_ids):
tweet = tweets[tid]['tweet']
parent = tweets[tid]['parent']
origin = tweets[tid]['origin']
root = tweets[tid]['origin_tweet']
cascade = tweets[tid]['cascade']
userid = tweets[tid]['user']
ptid = tweets[tid]['parent_tweet']
if cascade < 2:
continue
#bot filter
if bot.check_bot(Bot, userid) != 0:
continue
if userid in e_users:
e_child[tweets[tid]['depth']].append(tweets[tid]['child'])
else:
ne_child[tweets[tid]['depth']].append(tweets[tid]['child'])
if tweets[tid]['depth'] > 1:
diff = (parser.parse(tweets[tid]['time']) - parser.parse(
tweets[ptid]['time'])).total_seconds() / 60
if e_time[tweets[ptid]['depth']].get(ptid, -1) > diff:
print(e_time[tweets[ptid]['depth']][ptid], diff)
if parent in e_users:
# if origin in e_users:
if e_time[tweets[ptid]['depth']].get(ptid, -1) == -1:
e_time[tweets[ptid]['depth']][ptid] = diff
else:
if ne_time[tweets[ptid]['depth']].get(ptid, -1) == -1:
ne_time[tweets[ptid]['depth']][ptid] = diff
#if ccc == 5:
# break
#remove child 0 count
for i in range(1, 20):
e_child[i] = [x for x in e_child[i] if x != 0]
ne_child[i] = [x for x in ne_child[i] if x != 0]
box = BoxPlot(1)
box.set_multiple_data([e_child, ne_child])
box.set_ylog()
box.set_label('Depth', 'Child Count')
box.save_image('Image/%s/child_num_wo_propagation.png' % folder)
for i in range(1, 20):
e_time[i] = e_time[i].values()
ne_time[i] = ne_time[i].values()
ne_time2[i] = ne_time2[i].values()
#print(e_time)
#print(ne_time)
box = BoxPlot(1)
box.set_multiple_data([e_time, ne_time])
box.set_ylog()
box.set_label('Depth', 'Propagation Time')
box.save_image('Image/%s/child_time_propagation.png' % folder)
with open('Data/Figure/5_3_1.json', 'w') as f:
json.dump(
{
'e_time': e_time,
'ne_time': ne_time,
'e_child': e_child,
'ne_child': ne_child
}, f)
def diversity(filename):
index = filename.replace(".json", "").split('echo_chamber')
print(index)
with open(filename) as f:
echo_chambers = json.load(f)
print('total ', len(echo_chambers))
friends_cache = {}
postid = {}
count = 0
echo_diversity = {}
for key in echo_chambers:
#print(key)
users = echo_chambers[key]
#print(users)
count += 1
if count % 100 == 0:
print(count)
#break
if len(users) < 2:
continue
postids = key.split('_')
for k in postids:
postid[k] = 1
#print(len(users))
polars = []
user_count = 0
#polarity scores
for userid in users:
score = get_polarity(userid)
if score != -999:
polars.append(score)
user_count += 1
postid[postids[0]] = user_count
postid[postids[1]] = user_count
if 1 in postid.values():
break
diversity = util.eta(polars)
echo_diversity[key] = diversity
random_diversity = {}
for key in postid.keys():
#number of users
user_num = postid[key]
#print(user_num)
with open(dir_name + key, 'r') as f:
tweets = json.load(f)
users = [tweet['user']for tweet in tweets.values()]
users = get_random_user(users, user_num)
polars = []
#polarity scores
for userid in users:
score = get_polarity(userid)
if score != -999:
polars.append(score)
diversity = util.eta(polars)
random_diversity[key] = diversity
#print(users)
#print(polars)
#print(diversity)
with open('Data/echo_chamber_diversity.json', 'w') as f:
json.dump({'echo_chamber':echo_diversity, 'random':random_diversity}, f)
box = BoxPlot(1)
box.set_data([random_diversity.values(), echo_diversity.values()],'')
box.set_xticks(['Random', 'Echo chamber'])
box.save_image('Image/diversity_box_%s.png'%index[1])
def polarity_diversity():
#check rumor polarity similarity
#check cascade polarity similarity
files = os.listdir(dir_name)
users_polarity = {}
users_polarity_cascade = {}
retweet_cache = {}
for ccc, postid in enumerate(files):
users_polarity[postid] = {}
with open(dir_name+ '%s'%postid, 'r') as f:
tweets = json.load(f)
retweet_cache[postid] = tweets
for tweet in tweets.values():
p_score = get_polarity(tweet['user'])
users_polarity[postid][tweet['user']] = p_score
users_polarity_cascade[tweet['origin_tweet']] = users_polarity_cascade.get(tweet['origin_tweet'], {})
users_polarity_cascade[tweet['origin_tweet']][tweet['user']] = p_score
#if ccc == 10:
# break
r_diversity = []
for key in users_polarity.keys():
r_diversity.append(util.eta([score for score in users_polarity[key].values()]))
print(pd.Series(r_diversity).describe())
c_diversity = []
for key in users_polarity_cascade.keys():
if len(users_polarity_cascade[key]) < 2:
continue
c_diversity.append(util.eta([score for score in users_polarity_cascade[key].values()]))
print(pd.Series(c_diversity).describe())
box = BoxPlot(1)
box.set_data([r_diversity, c_diversity],'')
box.set_xticks(['Rumor', 'Cascade'])
box.save_image('Image/%s/diversity_box.png'%foldername)
#check echo chamber users' poarltiy similarity
e_diversity = []
echo_chamber_users = {}
with open('Data/echo_chamber2.json') as f:
echo_chamber = json.load(f)
for key in echo_chamber:
users = echo_chamber[key]
if len(users) < 2:
continue
polar = []
e_diversity.append(util.eta([get_polarity(user) for user in users]))
#get all echo chamber users for cascade characteristics
for postid in key.split('_'):
echo_chamber_users[postid] = echo_chamber_users.get(postid, {})
for user in users:
echo_chamber_users[postid][user] = 1
print(pd.Series(e_diversity).describe())
#check echo chamber user pariticpate polarity similarity and non-echo chamber user participate polarity similarity
echo_cascade = {}
cascade_users = {}
for postid in files:
tweets = retweet_cache[postid]
#get echo chamber cascade
for tweet in tweets.values():
if tweet['user'] in echo_chamber_users[postid].keys():
echo_cascade[tweet['origin_tweet']] = 1
cascade_users[tweet['origin_tweet']] = cascade_users.get(tweet['origin_tweet'], {})
cascade_users[tweet['origin_tweet']][tweet['user']] = get_polarity(tweet['user'])
echo_cascade = echo_cascade.keys()
echo_cascade_diversity = []
echo_cascade_size = []
non_echo_cascade_diversity = []
non_echo_cascade_size = []
for cascade in cascade_users.keys():
#echo chamber user participated cascade
if cascade in echo_cascade:
echo_cascade_diversity.append(utily.eta([score for score in cascade_users[cascade].values()]))
echo_cascade_size.append(len(cascade_users[cascade]))
#non echo chamber user participated cascade
else:
non_echo_cascade_diversity.append(util.eta([score for score in cascade_users[cascade].values()]))
non_echo_cascade_size.append(len(cascade_users[cascade]))
print('echo chamber cascade')
print(pd.Series(echo_cascade_diversity).describe())
print(pd.Series(echo_cascade_size).describe())
print('non echo chamber cascade')
print(pd.Series(non_echo_cascade_diversity).describe())
print(pd.Series(non_echo_cascade_size).describe())
box = BoxPlot(1)
box.set_data([echo_cascade_diversity, non_echo_cascade_diversity],'')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.save_image('Image/20180927/diversity_echo_cascade_box.png')
box = BoxPlot(1)
box.set_data([echo_cascade_size, non_echo_cascade_size],'')
box.set_xticks(['Echo Chamber', 'Non Echo Chamber'])
box.save_image('Image/20180927/diversity_echo_cascade_size_box.png')
def draw_5_3_1_figures():
with open('Data/Figure/5_3_1_2.json', 'r') as f:
data = json.load(f)
e_child = data['e_child']
ne_child = data['ne_child']
e_time = data['e_time']
ne_time = data['ne_time']
ne_time2 = data['ne_time2']
e_c = {}
e_t = {}
ne_c = {}
ne_t = {}
ne_t2 = {}
for i in range(1, 11):
e_c[i] = e_child[str(i)]
e_t[i] = e_time[str(i)]
ne_c[i] = ne_child[str(i)]
ne_t[i] = ne_time[str(i)]
ne_t2[i] = ne_time2[str(i)]
#if i > 6:
# print(e_t[i])
# print(e_c[i])
box = BoxPlot(1)
box.set_multiple_data([e_c, ne_c])
box.set_ylog()
box.set_label('Depth', 'Child Count')
box.save_image('Image/Figure/5_3_1_2.png')
print(e_t.keys())
box = BoxPlot(1)
box.set_multiple_data([e_t, ne_t])
box.set_ylog()
box.set_label('Depth', 'Propagation Time')
box.set_yticks(['0', '1 m', '5 m', '1 h', '1 day'],
index=[0, 1, 5, 60, 24 * 60])
#box.set_yticks(['0', '1 m', '10 m', '1 h', '1 day'], index=[0,1,10,60, 24*60])
box.save_image('Image/Figure/5_3_1_1.png')
#filter the wrong value is duration over 6 month
filter_value = 60 * 24 * 180
e_time['1'] = [item for item in e_time['1'] if item < filter_value]
ne_time['1'] = [item for item in ne_time['1'] if item < filter_value]
print(max(e_time['1']))
print(max(ne_time['1']))
e_time['1'] = sorted(e_time['1'])
ne_time['1'] = sorted(ne_time['1'])
#print(e_time['1'])
#print(ne_time['1'])
draw_cdf_plot([e_time['1'], ne_time['1']], 'Propagation Time',
['Echo chamber', 'Non-echo chamber'], '',
'Image/Figure/5_3_2.png')