协同过滤代码--getRating.py文件

#coding=utf-8


from math import sqrt
from loadMovieLens import loadMovieLensTrain
from loadMovieLens import loadMovieLensTest
    
### 计算pearson相关度
def sim_pearson(prefer, person1, person2):
    sim = {}
    #查找双方都评价过的项
    for item in prefer[person1]:
        if item in prefer[person2]:
            sim[item] = 1           #将相同项添加到字典sim中
    #元素个数
    n = len(sim)
    if len(sim)==0:
        return -1

    # 所有偏好之和
    sum1 = sum([prefer[person1][item] for item in sim])  
    sum2 = sum([prefer[person2][item] for item in sim])  

    #求平方和
    sum1Sq = sum( [pow(prefer[person1][item] ,2) for item in sim] )
    sum2Sq = sum( [pow(prefer[person2][item] ,2) for item in sim] )

    #求乘积之和 ∑XiYi
    sumMulti = sum([prefer[person1][item]*prefer[person2][item] for item in sim])

    num1 = sumMulti - (sum1*sum2/n)
    num2 = sqrt( (sum1Sq-pow(sum1,2)/n)*(sum2Sq-pow(sum2,2)/n))  
    if num2==0:                                                ### 如果分母为0，本处将返回0.
        return 0  

    result = num1/num2
    return result


### 获取对item评分的K个最相似用户（K默认20）
def topKMatches(prefer, person, itemId, k=20, sim = sim_pearson):
    userSet = []
    scores = []
    users = []
    #找出所有prefer中评价过Item的用户,存入userSet
    for user in prefer:
        if itemId in prefer[user]:
            userSet.append(user)
    #计算相似性
    scores = [(sim(prefer, person, other),other) for other in userSet if other!=person]

    #按相似度排序
    scores.sort()
    scores.reverse()

    if len(scores)<=k:       #如果小于k，只选择这些做推荐。
        for item in scores:
            users.append(item[1])  #提取每项的userId
        return users
    else:                   #如果>k,截取k个用户
        kscore = scores[0:k]
        for item in kscore:
            users.append(item[1])  #提取每项的userId
        return users               #返回K个最相似用户的ID


### 计算用户的平均评分
def getAverage(prefer, userId):
    count = 0
    sum = 0
    for item in prefer[userId]:
        sum = sum + prefer[userId][item]
        count = count+1
    return sum/count


### 平均加权策略，预测userId对itemId的评分
def getRating(prefer1, userId, itemId, knumber=20,similarity=sim_pearson):
    sim = 0.0
    averageOther =0.0
    jiaquanAverage = 0.0
    simSums = 0.0
    #获取K近邻用户(评过分的用户集)
    users = topKMatches(prefer1, userId, itemId, k=knumber, sim = sim_pearson)

    #获取userId 的平均值
    averageOfUser = getAverage(prefer1, userId)     

    #计算每个用户的加权，预测 
    for other in users:
        sim = similarity(prefer1, userId, other)    #计算比较其他用户的相似度
        averageOther = getAverage(prefer1, other)   #其他用户的平均分
        # 累加
        simSums += abs(sim)    #取绝对值
        jiaquanAverage +=  (prefer1[other][itemId]-averageOther)*sim   #累加，一些值为负

    # simSums为0，即该项目尚未被其他用户评分，这里的处理方法：返回用户平均分
    if simSums==0:
        return averageOfUser
    else:
        return (averageOfUser + jiaquanAverage/simSums)  


##==================================================================
##     getAllUserRating(): 获取所有用户的预测评分，存放到fileResult中
##
## 参数:fileTrain,fileTest 是训练文件和对应的测试文件，fileResult为结果文件
##     similarity是相似度度量方法，默认是皮尔森。
##==================================================================
def getAllUserRating(fileTrain='u1.base', fileTest='u1.test', fileResult='result.txt', similarity=sim_pearson):
    prefer1 = loadMovieLensTrain(fileTrain)         # 加载训练集 
    prefer2 = loadMovieLensTest(fileTest)           # 加载测试集  
    inAllnum = 0

    file = open(fileResult, 'a')
    file.write("%s\n"%("------------------------------------------------------"))
    
    for userid in prefer2:             #test集中每个用户
        for item in prefer2[userid]:   #对于test集合中每一个项目用base数据集,CF预测评分
            rating = getRating(prefer1, userid, item, 20)   #基于训练集预测用户评分(用户数目<=K)
            file.write('%s\t%s\t%s\n'%(userid, item, rating))
            inAllnum = inAllnum +1
    file.close()
    print("-------------Completed!!-----------",inAllnum)


############    主程序   ##############
if __name__ == "__main__":
    print("\n--------------推荐系统 运行中... -----------\n")
    getAllUserRating('u1.base', 'u1.test', 'result.txt')