#!/usr/bin/python
# coding=utf-8
from numpy import *
# 過(guò)濾網(wǎng)站的惡意留言 侮辱性：1   非侮辱性：0
# 創(chuàng)建一個(gè)實(shí)驗(yàn)樣本
def loadDataSet():
  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]
  return postingList, classVec
# 創(chuàng)建一個(gè)包含在所有文檔中出現(xiàn)的不重復(fù)詞的列表
def createVocabList(dataSet):
  vocabSet = set([])   # 創(chuàng)建一個(gè)空集
  for document in dataSet:
    vocabSet = vocabSet | set(document)  # 創(chuàng)建兩個(gè)集合的并集
  return list(vocabSet)
# 將文檔詞條轉(zhuǎn)換成詞向量
def setOfWords2Vec(vocabList, inputSet):
  returnVec = [0]*len(vocabList)    # 創(chuàng)建一個(gè)其中所含元素都為0的向量
  for word in inputSet:
    if word in vocabList:
      # returnVec[vocabList.index(word)] = 1   # index函數(shù)在字符串里找到字符第一次出現(xiàn)的位置 詞集模型
      returnVec[vocabList.index(word)] += 1   # 文檔的詞袋模型  每個(gè)單詞可以出現(xiàn)多次
    else: print "the word: %s is not in my Vocabulary!" % word
  return returnVec
# 樸素貝葉斯分類(lèi)器訓(xùn)練函數(shù)  從詞向量計(jì)算概率
def trainNB0(trainMatrix, trainCategory):
  numTrainDocs = len(trainMatrix)
  numWords = len(trainMatrix[0])
  pAbusive = sum(trainCategory)/float(numTrainDocs)
  # p0Num = zeros(numWords); p1Num = zeros(numWords)
  # p0Denom = 0.0; p1Denom = 0.0
  p0Num = ones(numWords);  # 避免一個(gè)概率值為0,最后的乘積也為0
  p1Num = ones(numWords);  # 用來(lái)統(tǒng)計(jì)兩類(lèi)數(shù)據(jù)中，各詞的詞頻
  p0Denom = 2.0; # 用于統(tǒng)計(jì)0類(lèi)中的總數(shù)
  p1Denom = 2.0 # 用于統(tǒng)計(jì)1類(lèi)中的總數(shù)
  for i in range(numTrainDocs):
    if trainCategory[i] == 1:
      p1Num += trainMatrix[i]
      p1Denom += sum(trainMatrix[i])
    else:
      p0Num += trainMatrix[i]
      p0Denom += sum(trainMatrix[i])
      # p1Vect = p1Num / p1Denom
      # p0Vect = p0Num / p0Denom
  p1Vect = log(p1Num / p1Denom)  # 在類(lèi)1中，每個(gè)次的發(fā)生概率
  p0Vect = log(p0Num / p0Denom)   # 避免下溢出或者浮點(diǎn)數(shù)舍入導(dǎo)致的錯(cuò)誤  下溢出是由太多很小的數(shù)相乘得到的
  return p0Vect, p1Vect, pAbusive
# 樸素貝葉斯分類(lèi)器
def classifyNB(vec2Classify, p0Vec, p1Vec, pClass1):
  p1 = sum(vec2Classify*p1Vec) + log(pClass1)
  p0 = sum(vec2Classify*p0Vec) + log(1.0-pClass1)
  if p1 > p0:
    return 1
  else:
    return 0
def testingNB():
  listOPosts, listClasses = loadDataSet()
  myVocabList = createVocabList(listOPosts)
  trainMat = []
  for postinDoc in listOPosts:
    trainMat.append(setOfWords2Vec(myVocabList, postinDoc))
  p0V, p1V, pAb = trainNB0(array(trainMat), array(listClasses))
  testEntry = ['love','my','dalmation']
  thisDoc = array(setOfWords2Vec(myVocabList, testEntry))
  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)
# 調(diào)用測(cè)試方法----------------------------------------------------------------------
testingNB()