similiar_users.py

from pymongo import MongoClient
import datetime

client = MongoClient('localhost',27017)
db = client['convflowdev']
problems = db['problems']
users = db['users']


def count_keywords(posted_problem_keywords,search_problem_keywords):
    count = 0
    for i in posted_problem_keywords:
        if i in search_problem_keywords:
            count = count + 1
    return count
    

def find_similiar_problems(keywords):
    """
    (list) --> objectIDs
    """
    list_of_ids = [(i['_id'],count_keywords(keywords,i['keywords'])) \
                   for i in problems.find({'keywords':{'$in':keywords}})]
##    print 'list_of_ids --> ' + str(list_of_ids) #ids are problems IDs
    return sorted(list_of_ids, key=lambda ids : ids[1],reverse=True)

def distribute_list_and_count_activity_index(id_list):

    l1 = [i[0] for i in id_list if i[1]>3]
    l2 = [i[0] for i in id_list if i[1]>1 and i[1]<4]
    l3 = [i[0] for i in id_list if i[1]==1]
    
##    print 'list l1 --> ' + str(l1)
##    print 'list l2 --> ' + str(l2)
##    print 'list l3 --> ' + str(l3)
##    print ''
##    print 'f_user_list --> ' + str(f_user_list)
##    print ''
    user_list_1 = []
    user_list_2 = []
    user_list_3 = []

##    print 'working on l1'
    for i in l1:
        #post = problems.find_one(
        a = problems.find_one({'_id':i})
##        print ''
##        print a
        user_list_1 = user_list_1 + a['acl']['users']

    ul_1 = {i:user_list_1.count(i) for i in set(user_list_1)}
##    print 'ul_1 --> ' + str(ul_1)
##    print ''
##    print 'working on l2'
    for i in l2:
        a = problems.find_one({'_id':i})
##        print ''
##        print a
        user_list_2 = user_list_2 + a['acl']['users']

    ul_2 = {i:user_list_2.count(i) for i in set(user_list_2)}
##    print 'ul_2 --> ' + str(ul_2)
##
##    print ''
##    print 'working on l3'
    for i in l3:
        a = problems.find_one({'_id':i})
##        print ''
##        print a
        user_list_3 = user_list_3 + a['acl']['users']

##    print ''
##    print 'user_list_1 --> ' + str(user_list_1)
##    print 'user_list_2 --> ' + str(user_list_2)
##    print 'user_list_3 --> ' + str(user_list_3)
##    print ''
    f_user_list = set(user_list_1 + user_list_2 + user_list_3)
    ul_3 = {i:user_list_3.count(i) for i in set(user_list_3)}
##    print 'ul_3 --> ' + str(ul_3)
    
    final_list = {}
    #net_scores = []
    for i in f_user_list:
        
        if i not in set(user_list_1):
            score_1 = 0
        else:
            score_1 = ul_1[i]

        if i not in set(user_list_2):
            score_2 = 0
        else:
            score_2 = ul_2[i]

        if i not in set(user_list_3):
            score_3 = 0
        else:
            score_3 = ul_3[i]

        net_score = score_1*0.6 + score_2*0.25 + score_3*0.15
        #net_scores.append(net_score)
        final_list[i] = net_score
##        print 'final_list[' + str(net_score)+'] = ' + str(i)
    ##        print ''
    all_scores = set(final_list.values())
    ff_dict = {}
    for i in all_scores:
        tmp_list = []
        for j in final_list:
            if final_list[j] == i:
                tmp_list.append(j)
        ff_dict[i] = tmp_list
    return ff_dict

def select_no_of_users(keywords,number):
    """

    integer -> list of first no. of 'integer' users with highest activity
    
    """
    final_list = distribute_list_and_count_activity_index(find_similiar_problems(keywords))
##    print 'final_list --> ' + str(final_list)
    netscores = sorted(final_list,reverse=True)
    
    required_scores = netscores[:number]
    req_users = []
    
    for i in required_scores:
        if(len(req_users)<number):
            req_users = req_users + final_list[i]
        else:
            break

    for i in req_users:
        a = users.find_one({'_id':i})
        print a['username']

    return req_users