def getPartialInfluenceScore():
clusterInfoCollection = getPagesClusterInfoCollection()
clusterInfluencerCollection = getClusterInfluencerScoreCollection()
friendsCollection = getFriendsCollection()
clusterInfluencerCollection.drop()
clusterNumber = len(clusterInfoCollection.distinct('cluster'))
for friend in friendsCollection.find():
_id = friend['id']
document = {}
document['_id'] = _id
document['cluster'] = {}
for i in range(clusterNumber):
document['cluster'][str(i)] = []
clusterInfluencerCollection.insert(document)
pagesCursor = clusterInfoCollection.find({"count": {"$gt": 3}})
epoch = datetime.datetime.utcfromtimestamp(0)
dt = 7 * 24 * 60 * 60
scores = []
done = 0
for page in pagesCursor:
users = page['people']
try:
users.sort(key=lambda x: x['created_time'])
cluster = page['cluster']
liketime = []
for user in users:
liketime.append((user['created_time'] - epoch).total_seconds())
back = 0
done += 1
print done
for user in users:
userId = user['id']
timeahead = (user['created_time'] - epoch).total_seconds() + dt
timeback = (user['created_time'] - epoch).total_seconds() - dt
ahead = bisect.bisect_right(liketime, timeahead)
score = ahead - back
back += 1
# print userId, cluster, score
clusterInfluencerCollection.update(
{'_id': userId},
{'$push': {
'cluster.' + str(cluster): score
}},
upsert=False)
except:
print "hmmm"
def getPartialInfluenceScore():
clusterInfoCollection = getPagesClusterInfoCollection()
clusterInfluencerCollection = getClusterInfluencerScoreCollection()
friendsCollection = getFriendsCollection()
clusterInfluencerCollection.drop()
clusterNumber = len(clusterInfoCollection.distinct('cluster'))
for friend in friendsCollection.find():
_id = friend['id']
document = {}
document['_id'] = _id
document['cluster'] = {}
for i in range(clusterNumber):
document['cluster'][str(i)] = []
clusterInfluencerCollection.insert(document)
pagesCursor = clusterInfoCollection.find({"count": {"$gt": 3}})
epoch = datetime.datetime.utcfromtimestamp(0)
dt = 7 * 24 * 60 * 60
scores = []
done = 0
for page in pagesCursor:
users = page['people']
try:
users.sort(key=lambda x: x['created_time'])
cluster = page['cluster']
liketime = []
for user in users:
liketime.append((user['created_time'] - epoch).total_seconds())
back = 0
done += 1
print done
for user in users:
userId = user['id']
timeahead = (user['created_time'] - epoch).total_seconds() + dt
timeback = (user['created_time'] - epoch).total_seconds() - dt
ahead = bisect.bisect_right(liketime, timeahead)
score = ahead - back
back += 1
# print userId, cluster, score
clusterInfluencerCollection.update({'_id': userId}, {'$push': {'cluster.' + str(cluster): score}},
upsert=False)
except:
print "hmmm"
from database import getPageCollection, getLikesCollection, getPagesClusterInfoCollection, getClusterCollection
from pprint import pprint
import dateutil.parser as dateparser
allpages = getPageCollection()
alllikes = getLikesCollection()
fbpagesinfo = getPagesClusterInfoCollection()
clusterinfo = getClusterCollection()
fbpagesinfo.drop()
counter = 0
for pageId in allpages.find():
cursor = alllikes.find({'data': {'$elemMatch': {'id': pageId['_id']}}})
cluster = clusterinfo.find_one({'pages': pageId['_id']})
cluster = cluster["cluster"]
document = {'_id': pageId['_id'], 'people': [], 'count': cursor.count(), 'cluster': cluster}
for c in cursor:
dd = {'id': c['id']}
for pages in c['data']:
if pages['id'] == pageId['_id']:
if 'created_time' in pages:
dd['created_time'] = dateparser.parse(pages['created_time'])
break
document['people'].append(dd)
counter += 1
print 'document', counter, 'done'
fbpagesinfo.insert(document)
from database import getPageCollection, getLikesCollection, getPagesClusterInfoCollection, getClusterCollection
from pprint import pprint
import dateutil.parser as dateparser
allpages = getPageCollection()
alllikes = getLikesCollection()
fbpagesinfo = getPagesClusterInfoCollection()
clusterinfo = getClusterCollection()
fbpagesinfo.drop()
counter = 0
for pageId in allpages.find():
cursor = alllikes.find({'data': {'$elemMatch': {'id': pageId['_id']}}})
cluster = clusterinfo.find_one({'pages': pageId['_id']})
cluster = cluster["cluster"]
document = {
'_id': pageId['_id'],
'people': [],
'count': cursor.count(),
'cluster': cluster
}
for c in cursor:
dd = {'id': c['id']}
for pages in c['data']:
if pages['id'] == pageId['_id']:
if 'created_time' in pages:
dd['created_time'] = dateparser.parse(
pages['created_time'])
break
document['people'].append(dd)
