这里假设 类标签为largeDoses, smallDoses, didntLike三类，假设训练样本有三个特征属性，类标签放在数据集的最后一列

import numpy as npdef file2matrix(filename):  # filename是文件保存地址love_dictionary = {'largeDoses':3, 'smallDoses':2, 'didntLike':1}fr = open(filename)arrayOLines = fr.readlines()numberOfLines = len(arrayOLines)            # 获得文件的行数returnMat = np.zeros((numberOfLines, 3))        # 用于存放训练数据classLabelVector = []                       # 用于存放类标签index = 0for line in arrayOLines:line = line.strip() # 截取掉所有的回车字符listFromLine = line.split() returnMat[index, :] = listFromLine[0:3] # 存放训练样本if(listFromLine[-1].isdigit()): # 如果标签字符串是数字，用int（）函数转换为数字类型classLabelVector.append(int(listFromLine[-1]))else: # 如果标签字符串不是数字，利用字典转换为数字类型classLabelVector.append(love_dictionary.get(listFromLine[-1]))index += 1return returnMat, classLabelVector

isdigit()判断一个字符串是否为数字

b = ['a', '2a', '2']
print(b[0].isdigit())
print(b[1].isdigit())
print(b[2].isdigit())False
False
True