• 技术文章 >常见问题 >Python常见问题

    在python中如何打乱数据?

    宋雪维宋雪维2020-12-03 17:43:58原创6289

    在玩python学习机器时,对于那种对随机性不太敏感的模型,理论上说可以不打乱。但敏感不敏感也跟数据量级,复杂度,算法内部计算机制都有关,目前并没有一个经纬分明的算法随机度敏感度列表。既然打乱数据并不会得到一个更差的结果,一般推荐的做法就是打乱全量数据。那怎么打乱呢?今天小编就教大家在python中打乱数据集和标签,来看看吧。

    方法一、打乱索引Index

    1

    2

    3

    4

    5

    6

    7

    8

    9

    import numpy as np

     

    index = [i for i in range(len(test_data))] # test_data为测试数据

     

    np.random.shuffle(index) # 打乱索引

     

    test_data = test_data[index]

     

    test_label = test_label[index]

    方法二:通过数组来shuffle来打乱

    1

    2

    3

    4

    5

    6

    7

    8

    9

    10

    11

    12

    13

    14

    15

    16

    17

    <p style="line-height: normal;"><span style="font-family: 微软雅黑,Microsoft YaHei;">image_list=[]           # list of images

     

    label_list=[]           # list of labels

     

      

     

    temp = np.array([image_list, label_list])

     

    temp = temp.transpose()

     

    np.random.shuffle(temp)

     

      

     

    images = temp[:, 0]     # array of images   (N,)

     

    labels = temp[:, 1]</span><br></p>

    方法三:通过随机数打乱

    1

    2

    3

    4

    5

    6

    7

    8

    9

    import numpy as np

     

    np.random.seed(12)

     

    np.random.shuffle(test_data)

     

    np.random.seed(12)

     

    np.random.shuffle(test_label)

    以上就是小编整理的用python打乱数据的方法,如果你在玩python机器学习的话,可以采取以上方法打乱数据哦~更多学习推荐:python学习网。














    专题推荐:python常见问题
    上一篇:python中的shuffle怎么用? 下一篇:如何用python发送邮件给多人?

    相关文章推荐

    • python入门:int()• 如何使用python3代码输出嵌套式对象?

    全部评论我要评论

    © 2021 Python学习网 苏ICP备2021003149号-1

  • 取消发布评论
  • 

    Python学习网