携程笔试大数据题，贝叶斯分类，一行代码，60%通过

更新一下ac了的第一题:

	



	def getH(arr):
    feature = [i[0] for i in arr]
    target = [i[1] for i in arr]
    num_0 = target.count(0)
    num_1 = target.count(1)
    length = len(target)
    if num_0 and num_1:
        H_t = -(num_0/length*log2(num_0/length)+num_1/length*log2(num_1/length))
    else:
        H_t = 0
    return H_t
def getGain(arr,n):
    H = getH(arr)
    feature = [i[0] for i in arr]
    target = [i[1] for i in arr]
    fs = [[i for i in l if i[0] == j] for j in set(feature)]
    g = 0
    for i in fs:       
        g += len(i)/n*getH(i)
    return H-g