决策树-选择最优的划分数据集的特点

def chooseBestFeatureToSplit(dataset):
    numFeature = len(dataset[0])-1   #划分后特点的数量
    BasehannonEnt= calcshannonEnt(dataset)  #初始的香农熵
    BestshannonEnt = 0.0    #初始化最优香农熵
    BestInfoGain = 0.0    #最优信息增益
    BestFeature = -1   #最优特点的坐标
    for i in range(numFeature):
        featList = [sample[i] for sample in dataset]   #提取出特点在数据集中的取值
        uniqueVals = set(featList)     #找到可能的取值
        newEntropy = 0.0    #初始化新香农熵
        for value in uniqueVals:   #遍历每种可能划分数据集，因为有两个除数据集以外的参数，所以两重循环
            subdataset = splitdataset(dataset,i,value)  
            prob = len(subdataset)/float(len(dataset))
            newEntropy  +=prob *calcshannonEnt(subdataset)  #分别计算他们的香农熵
            
        infoGain = BasehannonEnt - newEntropy   #信息增益
    #找到最优的信息增益
        if (infoGain>BestInfoGain):
            BestInfoGain = infoGain       
            BestFeature = i
    return BestFeature