标签:
0_5.txt
00000000000000110000000000000000
00000000000011111100000000000000
00000000000111111110000000000000
00000000001111111111000000000000
00000000111111111111100000000000
00000001111111111111110000000000
00000000111111111111111000000000
00000000111111100001111100000000
00000001111111000001111100000000
00000011111100000000111100000000
00000011111100000000111110000000
00000011111100000000011110000000
00000011111100000000011110000000
00000001111110000000001111000000
00000011111110000000001111000000
00000011111100000000001111000000
00000001111100000000001111000000
00000011111100000000001111000000
00000001111100000000001111000000
00000001111100000000011111000000
00000000111110000000001111100000
00000000111110000000001111100000
00000000111110000000001111100000
00000000111110000000011111000000
00000000111110000000111111000000
00000000111111000001111110000000
00000000011111111111111110000000
00000000001111111111111110000000
00000000001111111111111110000000
00000000000111111111111000000000
00000000000011111111110000000000
00000000000000111111000000000000
#-*- coding: utf-8 -*-
‘‘‘
Created on Mar 8, 2011
@author: Peter
‘‘‘
from numpy import *
from numpy import linalg as la
#输入数据:用户-菜肴
#行:用户,列:菜肴
#数字表示用户对菜肴的喜好评分1-5,0表示未品尝该菜
def loadExData0():
return[[0, 0, 0, 2, 2],
[0, 0, 0, 3, 3],
[0, 0, 0, 1, 1],
[1, 1, 1, 0, 0],
[2, 2, 2, 0, 0],
[5, 5, 5, 0, 0],
[1, 1, 1, 0, 0]]
def loadExData():
return[[4, 4, 0, 2, 2],
[4, 0, 0, 3, 3],
[4, 0, 0, 1, 1],
[1, 1, 1, 2, 0],
[2, 2, 2, 0, 0],
[1, 1, 1, 0, 0],
[5, 5, 5, 0, 0]]
def loadExData2():
return[[0, 0, 0, 0, 0, 4, 0, 0, 0, 0, 5],
[0, 0, 0, 3, 0, 4, 0, 0, 0, 0, 3],
[0, 0, 0, 0, 4, 0, 0, 1, 0, 4, 0],
[3, 3, 4, 0, 0, 0, 0, 2, 2, 0, 0],
[5, 4, 5, 0, 0, 0, 0, 5, 5, 0, 0],
[0, 0, 0, 0, 5, 0, 1, 0, 0, 5, 0],
[4, 3, 4, 0, 0, 0, 0, 5, 5, 0, 1],
[0, 0, 0, 4, 0, 4, 0, 0, 0, 0, 4],
[0, 0, 0, 2, 0, 2, 5, 0, 0, 1, 2],
[0, 0, 0, 0, 5, 0, 0, 0, 0, 4, 0],
[1, 0, 0, 0, 0, 0, 0, 1, 2, 0, 0]]
def loadExData1():
return[[2, 0, 0, 4, 4, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 5],
[0, 0, 0, 0, 0, 0, 0, 1, 0, 4, 0],
[3, 3, 4, 0, 3, 0, 0, 2, 2, 0, 0],
[5, 5, 5, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 5, 0, 0, 5, 0],
[4, 0, 4, 0, 0, 0, 0, 0, 0, 0, 5],
[0, 0, 0, 0, 0, 4, 0, 0, 0, 0, 4],
[0, 0, 0, 0, 0, 0, 5, 0, 0, 5, 0],
[0, 0, 0, 3, 0, 0, 0, 0, 4, 5, 0],
[1, 1, 2, 1, 1, 2, 1, 0, 4, 5, 0]]
#基于欧氏距离的相似度:1/(1+欧氏距离)
def ecludSim(inA,inB): #A,B两个样本向量
#库函数la.norm()计算欧式距离
return 1.0/(1.0 + la.norm(inA - inB))
#基于皮尔森相关系数的相似度:0.5+0.5*corrcoef()
def pearsSim(inA,inB):
if len(inA) < 3 : return 1.0
#corrcoef()返回的是2*2的对称矩阵,索引[0][1]或者[1][0]都行
#并将值范围[-1,1]归一化到[0,1]
return 0.5+0.5*corrcoef(inA, inB, rowvar = 0)[0][1]
#基于余弦夹角的相似度
def cosSim(inA,inB):
num = float(inA.T*inB)
denom = la.norm(inA)*la.norm(inB)
#将余弦夹角值[-1,1]归一化到[0,1]
return 0.5+0.5*(num/denom)
#基于物品相似度,计算用户对物品的评分估计值
#输入:dataMat 用户数据
#user:用户编号(行)
#simMeas:相似度计算函数
#item:物品编号(列),用户待预测的物品
def standEst(dataMat, user, simMeas, item):
#数据矩阵列,即为物品数
n = shape(dataMat)[1]
simTotal = 0.0; ratSimTotal = 0.0
#遍历所有物品
for j in range(n):
#用户user对j物品评分
userRating = dataMat[user,j]
#(1)若未对物品j评分,即userRating=0,不处理
if userRating == 0: continue
#(2)若对物品j评分:
#统计对物品item和物品j都评分的用户编号
overLap = nonzero(logical_and(dataMat[:,item].A>0, dataMat[:,j].A>0))[0]
#(2.1)若没有用户同时对物品item和j评分,则两物品间相似度为0
if len(overLap) == 0: similarity = 0
#(2.2)若有用户同时对物品item和j评分,抽取出来,计算相似度
else: similarity = simMeas(dataMat[overLap,item],dataMat[overLap,j])
print ‘the %d and %d similarity is: %f‘ % (item, j, similarity)
#相似度求和
simTotal += similarity
#预测用户user对物品item评分总和
ratSimTotal += similarity * userRating
if simTotal == 0: return 0
#归一化预测评分
else: return ratSimTotal/simTotal
#基于SVD的评分估计
#输入:dataMat 用户数据
#user:用户编号(行)
#simMeas:相似度计算函数
#item:物品编号(列),用户待预测的物品
def svdEst(dataMat, user, simMeas, item):
#物品数
n = shape(dataMat)[1]
simTotal = 0.0; ratSimTotal = 0.0
#SVD分解
U,Sigma,VT = la.svd(dataMat)
#构建对角矩阵,取前3个奇异值
#3个额外算出来的,确保总能量>90%
Sig3 = mat(eye(3)*Sigma[:3])
#SVD降维,重构低维空间的物品#.I求逆
xformedItems = dataMat.T * U[:,:3] * Sig3.I
#遍历所有物品
for j in range(n):
#用户user对j物品评分
userRating = dataMat[user,j]
#若未对物品j评分,即userRating=0,不处理
if userRating == 0 or j==item: continue
#在低维空间计算物品j与物品item的相似度
similarity = simMeas(xformedItems[item,:].T, xformedItems[j,:].T)
print ‘the %d and %d similarity is: %f‘ % (item, j, similarity)
#相似度求和
simTotal += similarity
#预测用户user对物品item评分总和
ratSimTotal += similarity * userRating
if simTotal == 0: return 0
#归一化预测评分
else: return ratSimTotal/simTotal
#基于物品相似度的推荐
#dataMat: 数据
#user: 用户编号
# N: 选择预测评分最高的N个结果
#simMeas: 相似度计算方法
#estMethod:用户对物品的预测估分方法
def recommend(dataMat, user, N=3, simMeas=cosSim, estMethod=standEst):
#找没有被用户user评分的物品
unratedItems = nonzero(dataMat[user,:].A==0)[1]
#若都评分则退出,不需要再推荐
if len(unratedItems) == 0: return ‘you rated everything‘
itemScores = []
#遍历未评分的物品
for item in unratedItems:
#预测用户user对为评分物品item的估分
estimatedScore = estMethod(dataMat, user, simMeas, item)
#存(物品编号,对应估分值)
itemScores.append((item, estimatedScore))
#选择最高的估分结果
return sorted(itemScores, key=lambda jj: jj[1], reverse=True)[:N]
#*****图像输出
def printMat(inMat, thresh=0.8):
for i in range(32):
for k in range(32):
#阈值>0.8,输出1.
if float(inMat[i,k]) > thresh:
print 1,
#阈值<0.8,输出0.
else: print 0,
print ‘‘
#*****图像压缩
def imgCompress(numSV=3, thresh=0.8):
myl = []
#读.txt文件,转换为矩阵存储到myMat中
for line in open(‘0_5.txt‘).readlines():
newRow = []
for i in range(32):
newRow.append(int(line[i]))
myl.append(newRow)
myMat = mat(myl)
print "****original matrix******"
#输出阈值处理的图像
printMat(myMat, thresh)
#SVD分解
U,Sigma,VT = la.svd(myMat)
#初始化numSV*numSV的零矩阵SigRecon
SigRecon = mat(zeros((numSV, numSV)))
#Sigma对角线前numSV个值重构对角矩阵SigRecon
for k in range(numSV):
SigRecon[k,k] = Sigma[k]
#重构后的图像矩阵
reconMat = U[:,:numSV]*SigRecon*VT[:numSV,:]
print "****reconstructed matrix using %d singular values******" % numSV
#输出阈值处理后的重构图像
printMat(reconMat, thresh)
if __name__=="__main__":
imgCompress(2)
# #(1)*****测试奇异值分解的结果:
# data=[[1,1,1,0,0],
# [2,2,2,0,0],
# [1,1,1,0,0],
# [5,5,5,0,0],
# [1,1,0,2,2],
# [0,0,0,3,3],
# [0,0,0,1,1]]
# U,sigma,VT=linalg.svd(data)
# print‘sigma=‘,sigma
# print‘U=‘,U
# print‘VT=‘,VT
# # #低秩重构,取前3个奇异值
# sigma3=mat([[sigma[0],0,0],[0,sigma[1],0],[0,0,sigma[2]]])
# recondata=U[:,:3]*sigma3*VT[:3,:]
# print‘recondata=‘
# print recondata
## #(2)*****基于物品相似度的推荐,未用到SVD
## data=loadExData()
## #data=loadExData2() #用的是loadExData2()
## data=mat(data)
## itemscores=recommend(data,5) #对用户2推荐
## print ‘recommend result:‘
## print itemscores
# #(3)*****svd 查看能量分布
# data=loadExData2()
# u,sigma,vT=linalg.svd(mat(data))
# print ‘sigma=‘, sigma
# sigma2=sigma**2
# print ‘all energy=‘,sum(sigma2)#总能量
# print ‘90% of all energy=‘,sum(sigma2)*0.9 #总能量的90%
# #sigma_len=len(sigma) #奇异值个数
# # for i in range(sigma_len)
# print ‘energy of the first 2 Singular value=‘,sum(sigma2[:2])#前几个奇异值能量求和
# print ‘energy of the first 3 Singular value=‘,sum(sigma2[:3])
#
#
# ##(4)*****基于svd推荐:
# data=loadExData2()
# data=mat(data)
# itemscores=recommend(data,1,estMethod=svdEst) #对用户1推荐,余弦夹角
# #对用户1推荐,皮尔逊系数
# #itemscores=svdRec.recommend(data,1,estMethod=svdRec.svdEst,simMeas=svdRec.pearsSim)
# print ‘recommend result:‘
# print itemscores
#-*- coding: utf-8 -*-
from numpy import*
from numpy import linalg as la
import svdRec
#图像SVD压缩
svdRec.imgCompress(2)
###*****SVD的调用,及sigma输出格式,如何取对角矩阵
##data=[[1, 1, 1, 0, 0],
## [2, 2, 2, 0, 0],
## [1, 1, 1, 0, 0],
## [5, 5, 5, 0, 0],
## [1, 1, 0, 2, 2],
## [0, 0, 0, 3, 3],
## [0, 0, 0, 1, 1]]
##
##u,sigma,vT=linalg.svd(data)
##print sigma
##sig3=mat([[sigma[0],0,0],[0,sigma[1],0],[0,0,sigma[2]]])
##recondata=u[:,:3]*sig3*vT[:3,:]
##print ‘recondata=‘,recondata
##
###*****基于物品相似度的推荐,未用到SVD
##data=svdRec.loadExData() #用的是loadExData()
###data=svdRec.loadExData2() #用的是loadExData2()
##data=mat(data)
##itemscores=svdRec.recommend(data,2) #对用户2推荐
##print ‘recommend result:‘
##print itemscores
###*****svd 查看能量分布
##data=svdRec.loadExData2()
##u,sigma,vT=linalg.svd(mat(data))
##print ‘sigma=‘, sigma
##sigma2=sigma**2
##print ‘all energy=‘,sum(sigma2)#总能量
##print ‘90% of all energy=‘,sum(sigma2)*0.9 #总能量的90%
##print ‘energy of the first 2 Singular value=‘,sum(sigma2[:2])#前几个奇异值能量求和
##print ‘energy of the first 3 Singular value=‘,sum(sigma2[:3])
##
####*****基于svd推荐:
##data=svdRec.loadExData2()
##data=mat(data)
##itemscores=svdRec.recommend(data,1,estMethod=svdRec.svdEst) #对用户1推荐,余弦夹角
###对用户1推荐,皮尔逊系数
###itemscores=svdRec.recommend(data,1,estMethod=svdRec.svdEst,simMeas=svdRec.pearsSim)
##print ‘recommend result:‘
##print itemscores
标签:
原文地址:http://blog.csdn.net/mlljava1111/article/details/50847500