Recommend.py

import numpy as np
import pandas as pd
from numpy import linalg as la
import random

# 读入数据的存储方式为行表示用户列表示电影值表示用户评分
# 由于用户数大于电影数，因此我们选择基于电影的相似度做推荐

data = pd.read_csv("data.csv")
user_score = data[['user_id', 'user_score', 'id']]
movies = set()
users = set()
user_index = {}
movie_index = {}
movie_reindex={}
global usernum
global movienum


# 计算相似度
def cosSim(inA, inB):  # inA inB两个列向量
    num = float(inA.T * inB)
    denom = la.norm(inA) * la.norm(inB)  # linalg.norm()计算范数
    return 0.5 + 0.5 * (num / denom)


def SVD(dataMat):  # 压缩矩阵dataMat
    dataMat = np.mat(dataMat)
    U, Sigma, VT = la.svd(dataMat)
    Sig4 = np.mat(np.eye(4) * Sigma[:4])
    # print(dataMat.T.shape,U[:,:4].shape,Sig4.I.shape)
    xformedItems = dataMat.T * U[:, :4] * Sig4.I
    return xformedItems


def preprocess():
    global usernum
    global movienum
    s = user_score['user_id'].value_counts()[:1000]
    usernum = len(s)
    users = s.index
    nu = 0
    for x in users:
        user_index[x] = nu
        nu += 1
    m = user_score['id'].value_counts()[:]
    movies = m.index
    movienum = len(movies)
    nm = 0
    for x in movies:
        movie_index[x] = nm
        movie_reindex[nm]=x
        nm += 1
    mat = np.zeros((usernum, movienum))
    for i in range(len(data)):
        cu = data.loc[i, 'user_id']
        cm = data.loc[i, 'id']
        if cu in users:
            if cm in movies:
                mat[user_index[cu], movie_index[cm]] = data.loc[i, 'user_score']
    # print(mat)
    return mat


def get_sim():
    global usernum #1000
    global movienum #
    SIM=np.zeros((movienum,movienum))
    for i in range(movienum):
        for j in range(movienum):
            SIM[i][j]=cosSim(xformedItems[i,:].T,xformedItems[j,:].T)
    return SIM

# def svdEst(dataMat,user,simMeas,item,SIM):
#     simTotal=0.0
#     ratSimTotal=0.0
#     for i in range(movienum):
#         # j=movie_index[i]
#         userRating=dataMat[user,i]
#         if userRating==0 or i==item: continue
#         similarity=SIM[i,item]
#         ratSimTotal+=similarity*userRating
#     if simTotal==0:
#         return 0
#     else:
#         return ratSimTotal/simTotal

def svdEst(dataMat,user,simMeas,item):
    n=np.shape(dataMat)[1]
    simTotal=0.0;ratSimTotal=0.0
    U,Sigma,VT=la.svd(dataMat)
    Sig4=np.mat(np.eye(4)*Sigma[:4])
    xformedItems=dataMat.T*U[:,:4]*Sig4.I
    for j in range(n):
        userRating=dataMat[user,j]
        if userRating==0 or j==item:continue
        similarity=simMeas(xformedItems[item,:].T,xformedItems[j,:].T)
        # print("the %d and %d similarity is: %f"%(item,j,similarity))
        simTotal+=similarity
        ratSimTotal+=similarity*userRating
    if simTotal==0:return 0
    else:
        return ratSimTotal/simTotal



def recommend(SIM,dataMat,user,simMeas=cosSim,estMethod=svdEst):
    dataMat=np.mat(dataMat)
    unratedItems=np.nonzero(dataMat[user,:]==0)[1]
    itemScores=[]
    for item in unratedItems:
        estimatedScore=estMethod(dataMat,user,simMeas,item)
        itemScores.append((item,estimatedScore))
    return sorted(itemScores,key=lambda x:x[1],reverse=True)
if __name__ == '__main__':
    dataMat = preprocess()
    xformedItems=SVD(dataMat)
    print("SVD has been done.")
    SIM=get_sim()
    print("SIM matrix has been done.")
    pointed_user=random.randint(0,usernum-1)
    REC=recommend(SIM,dataMat,pointed_user)
    result=[]
    for i in range(0,5):
        result.append(movie_reindex[REC[i][0]])
    print(result)# 这里的result就已经是推荐电影的结果 输出的是电影的ID
    #打印电影信息让你具体了解一下！
    result=iter(result)
    info=[]
    cur = next(result)
    while len(info)<5:
        for i in range(len(data)):
            if data.loc[i,'id']==cur:
                info.append(data.loc[i,['id','title','year','director','writer','actor','type',
                                    'country','language','length','score']])
                try:
                    cur = next(result)
                except:
                    break
                break
    print(info)