决策树

最新推荐文章于 2021-10-27 22:27:20 发布

Car12

最新推荐文章于 2021-10-27 22:27:20 发布

阅读量278

点赞数

分类专栏： Python

本文链接：https://blog.youkuaiyun.com/u010261063/article/details/79773668

版权

Python 专栏收录该内容

41 篇文章

订阅专栏

from math import log
import operator
def calcShannonEnt(dataSet):
    numEntries =len(dataSet)
    labelCount = {}
    for featVoc in dataSet:
        currentlabel = featVoc[-1]
        if currentlabel not in labelCount.keys():
            labelCount[currentlabel]=0
        labelCount[currentlabel] +=1
    shannonEnt = 0.0
    for key in labelCount:
        prob =float(labelCount[key])/numEntries
        # print(prob)
        shannonEnt -=prob*log(prob,2)
    return shannonEnt

def CreateDataSet():
    dataSet = [[1,1,'yes'],
               [1, 1, 'yes'],
               [1,0,'no'],
               [1,1,'no'],
               [1,0,'no']]
    label =['no surfacing','flippers']
    print(dataSet,label)
    return  dataSet,label

#                                                                                       [2,4,3]                               | 2 4|
def splitDataSet(dataSet,axis,value):   #抽取一个矩阵的指定列与value相等除外的其他列数据[1,2,2]    splitDataSet(data,1,3)==>  |    |
                                                                                       # [4,5,3]                              |4,6 |
    reData=[]
    for featVec in dataSet:
        if featVec[axis] == value:
            reduceFeatVec = featVec[:axis]
            reduceFeatVec.extend(featVec[axis+1:])
            # print(featVec[:axis],"---",featVec[axis+1:])
            # print(reduceFeatVec)
            reData.append(reduceFeatVec)
    return reData


def chaooseBestFeatureTosplit(dataSet):
    num = len(dataSet[0])-1
    baseEntropy = calcShannonEnt(dataSet)
    bestinfoGain = 0.0
    beseFeature = -1
    for i in range(num):
        featList = [x[i] for x in dataSet]
        # print(featList)
        uniqueVals = set(featList)
        # print("uniquesVals = ",uniqueVals)
        newEntropy=0.0
        for j in uniqueVals:
            subDataSet = splitDataSet(dataSet,i,j)
            prob = len(subDataSet)/len(dataSet)
            newEntropy += prob*calcShannonEnt(subDataSet)
        infoGain = baseEntropy-newEntropy
        # print(infoGain)
        if infoGain > bestinfoGain:
            beseFeature = i
            bestinfoGain = infoGain
    return beseFeature

def majoritycnt(classList):
    classCount = {}
    for vote in classList:
        if vote not in classCount.keys():
            classCount[vote]=0
        classCount[vote] +=1
        sortedClassCount = sorted(classCount.items(),key = operator.itemgetter(1),reverse=True)
        return sortedClassCount
def CreateTree(dataSet,labes):
    print(dataSet,'label ',labes)
    classList = [x[-1] for x in dataSet]
    cnt = len(set(classList))# classList.count(classList[0])  集合不能有重复元素
    # print(cnt,"   ",len(classList))
    # print(len(classList[-1]))
    if classList.count(classList[0])== len(classList): #除开最后一列，剩下的矩阵中第一列元素不相同数 == 行数
        print("aaaaa")
        return  classList[0]
    # print(len(dataSet[0]))
    if len(dataSet[0])==1:
        print("bbbbb")
        return majoritycnt(classList)
    bestFeature = chaooseBestFeatureTosplit(dataSet)
    bestFeatureLabe = labes[bestFeature]
    myTree = {bestFeatureLabe:{}}
    del labes[bestFeature]
    featValues = [x[bestFeature] for x in dataSet]
    uniqueVals  = set(featValues)
    # print(uniqueVals)

    # print(splitData)
    for value in uniqueVals:
        subLabel = labes[:]
        # print(subLabel)
        # print('-------',myTree)
        myTree[bestFeatureLabe][value]=CreateTree(splitDataSet(dataSet, bestFeature, value),subLabel)
        # print(myTree)
    return myTree



dataSet,labels = CreateDataSet();
# print(chaooseBestFeatureTosplit(dataSet))
print(CreateTree(dataSet, labels))