UserCF.py

import math
from operator import itemgetter, attrgetter
from collections import defaultdict
def UserSimilarity(train):
    #build inverse table for item_users
    print "build inverse table for item_users"
    item_users = defaultdict(set)
    for u,items in train.iteritems():
        for i in items.keys():
            item_users[i].add(u)
    
    #calculate co-rated items between users
    print 'calculate co-rated items between users'
    C = defaultdict(dict)
    N = defaultdict(int)
    for i,users in item_users.iteritems():
        for u in users:
            N[u] += 1
            for v in users:
                if u == v:
                    continue
                if v not in C[u]:
                    C[u][v] = 0
                C[u][v] += 1
                
    #calculate finial similarity matrix W
    print 'calculate finial similarity matrix W'
    W = defaultdict(dict)
    for u,related_users in C.items():
        for v,cuv in related_users.items():
            W[u][v] = cuv / math.sqrt(N[u] * N[v])
    
    return W
    
def LoadRating(trainFile):
    print 'loading rating'
    train = defaultdict(dict)
    with open(trainFile,'r') as f:
        for line in f:
            values = line.split('\t')
            uid = int(values[0])
            iid = int(values[1])
            rating = int(values[2])
            train[uid][iid] = rating
    
    return train