机器学习-朴素贝叶斯分类代码详解

from numpy import *

def loadDataSet():#positionlist相当于多个文档,每行为一个文档,classvec相当于他的标签
    postingList=[['my', 'dog', 'has', 'flea', 'problems', 'help', 'please'],
                 ['maybe', 'not', 'take', 'him', 'to', 'dog', 'park', 'stupid'],
                 ['my', 'dalmation', 'is', 'so', 'cute', 'I', 'love', 'him'],
                 ['stop', 'posting', 'stupid', 'worthless', 'garbage'],
                 ['mr', 'licks', 'ate', 'my', 'steak', 'how', 'to', 'stop', 'him'],
                 ['quit', 'buying', 'worthless', 'dog', 'food', 'stupid']]
    classVec = [0,1,0,1,0,1]    #1 is abusive, 0 not
    return postingList,classVec
                 
def createVocabList(dataSet):#将所有出现在dataset中的数据转换为不重复的单词列表,该列表包含了所有文档中出现的单词
    vocabSet = set([])  #create empty set
    for document in dataSet:
        vocabSet = vocabSet | set(document) #union of the two sets
    return list(vocabSet)

def setOfWords2Vec(vocabList, inputSet):#vocablist是输入的单词列表，inputset是单词匹配区，如果在vocablist中存在inputset中的单词，则在列表该位置值为1
    returnVec = [0]*len(vocabList)#列表*一个常数，则列表扩宽原来的常数被，内部的值重复,print([1,2,3,4,5]*3)>>[1, 2, 3, 4, 5, 1, 2, 3, 4, 5, 1, 2, 3, 4, 5]
    for word in inputSet:
        if word in vocabList:
            returnVec[vocabList.index(word)] = 1
        else: print ("the word: %s is not in my Vocabulary!" % word)#在最后的测试阶段,有可能拿到的文档信息中具有之前没有出现过的单词the word: dalmation1 is not in my Vocabulary!
    return returnVec#这里返回的是inputset中的数据,可以认为是某一个文档中的数据inputset,在所有文档中文字vocablist下出现的情况

def trainNB0(trainMatrix,trainCategory):#trainmatrix为文档矩阵,经上方的returnvec得来，traincategory为每篇文档类别标签
    numTrainDocs = len(trainMatrix)#矩阵的行为文档的个数
    numWords = len(trainMatrix[0])#矩阵的列为每一篇文档的字数
    pAbusive = sum(trainCategory)/float(numTrainDocs)#traincategory列表中的求和值,是列表中1的个数,numtraindocs是文档的个数,也是1类型文档占总文档的比例
    p0Num = ones(numWords); p1Num = ones(numWords)      #change to ones() ,numwords个元素的全1矩阵
    p0Denom = 2.0; p1Denom = 2.0                        #change to 2.0
    for i in range(numTrainDocs):
        if trainCategory[i] == 1:#当文档标签的类型为一时
            p1Num += trainMatrix[i]#将标签为1的文档中所有文字出现的次数进行纵向统计,因为在trainmatrix中的一行记录的是dataset的一行在所有文字中出现的情况
            p1Denom += sum(trainMatrix[i])#将标签为1的文档中,trainmatrix的每一行文字的字数求和,是1类型的总字数
        else:
            p0Num += trainMatrix[i]
            p0Denom += sum(trainMatrix[i])
    p1Vect = log(p1Num/p1Denom)          #change to log(),计算结果是在标签是1的所有文档下,每种文字在1类型总字数下的分别比值
    p0Vect = log(p0Num/p0Denom)          #change to log().计算结果是在标签是0的所有文档下,每种文字在0类型总字数下的分别比值
    #print(p1Num,p1Denom)
    return p0Vect,p1Vect,pAbusive#返回的结果是在文档标签类型为1和0的条件下各种文字出现的概率,pabusive是1类型文档占总文档的比例

def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):
    p1 = sum(vec2Classify * p1Vec) + log(pClass1)    #element-wise mult,两个向量对应位置的元素乘积
    p0 = sum(vec2Classify * p0Vec) + log(1.0 - pClass1)
    if p1 > p0:
        return 1
    else: 
        return 0
    
def bagOfWords2VecMN(vocabList, inputSet):
    returnVec = [0]*len(vocabList)
    for word in inputSet:
        if word in vocabList:
            returnVec[vocabList.index(word)] += 1
    return returnVec

def testingNB():
    listOPosts,listClasses = loadDataSet()#listOPosts含有文字的文档矩阵,listClasses文档矩阵每一行的标签属性
    myVocabList = createVocabList(listOPosts)##将所有出现在listOPosts中的数据转换为不重复的单词列表,该列表包含了所有文档中出现的单词
    trainMat=[]#空的文档矩阵,用来存放文档中各个文字出现的情况
    for postinDoc in listOPosts:
        trainMat.append(setOfWords2Vec(myVocabList, postinDoc))#使用append最后的结果为列表中的列表,有值的话对应位置结果为1,否则结果是0
    p0V,p1V,pAb = trainNB0(array(trainMat),array(listClasses))#计算结果是在标签是0和1的所有文档下,每种文字在0和1类型总字数下的分别比值
    testEntry = ['love', 'my', 'dalmation1']
    thisDoc = array(setOfWords2Vec(myVocabList, testEntry))#转换为和trainmat一样的形式
    print( testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb))
    testEntry = ['stupid', 'garbage']
    thisDoc = array(setOfWords2Vec(myVocabList, testEntry))
    print (testEntry,'classified as: ',classifyNB(thisDoc,p0V,p1V,pAb))

def textParse(bigString):    #input is big string, #output is word list
    import re
    listOfTokens = re.split(r'\W*', bigString)#http://blog.youkuaiyun.com/manjhok/article/details/78586818;以非字母字符进行分割,正则表达式
    return [tok.lower() for tok in listOfTokens if len(tok) > 2]

def spamTest():
    docList=[]; classList = []; fullText =[]
    for i in range(1,26):
        wordList = textParse(open('email/spam/%d.txt' % i).read())#wordlist是一个长度大于2的单词列表
        docList.append(wordList)#doclist是包含了单词列表的多维矩阵
        fullText.extend(wordList)#fulltext是一个包含所有单词包括重复了的单词列表
        classList.append(1)#添加分类标签1
        wordList = textParse(open('email/ham/%d.txt' % i).read())#读取非垃圾邮件的单词列表,一行垃圾邮件单词列表,一行正常邮件单词列表
        docList.append(wordList)
        fullText.extend(wordList)
        classList.append(0)#添加分类标签0
    vocabList = createVocabList(docList)#create vocabulary,#将所有出现在doclist中的数据转换为不重复的单词列表,该列表包含了所有文档中出现的单词,包括0类和1类
    trainingSet = range(50); testSet=[]           #create test set,垃圾邮件和正常邮件数目为26+26=52
    for i in range(10):
        randIndex = int(random.uniform(0,len(trainingSet)))
        testSet.append(trainingSet[randIndex])
        del(trainingSet[randIndex])  #删除之后这样的目的在于不会出现重复的结果，比如原来的6被删除，那么列表中就会只有，，4，5，7，，，
    trainMat=[]; trainClasses = []
    for docIndex in trainingSet:#train the classifier (get probs) trainNB0
        trainMat.append(bagOfWords2VecMN(vocabList, docList[docIndex]))#vocablist是所有单词出现情况的列表
        trainClasses.append(classList[docIndex])
    p0V,p1V,pSpam = trainNB0(array(trainMat),array(trainClasses))
    errorCount = 0
    for docIndex in testSet:        #classify the remaining items
        wordVector = bagOfWords2VecMN(vocabList, docList[docIndex])
        if classifyNB(array(wordVector),p0V,p1V,pSpam) != classList[docIndex]:
            errorCount += 1
            print ("classification error",docList[docIndex])
    print ('the error rate is: ',float(errorCount)/len(testSet))
    #return vocabList,fullText