python numpy数组和one-hot编码相互转换

最新推荐文章于 2025-03-24 18:28:19 发布

姚贤贤

最新推荐文章于 2025-03-24 18:28:19 发布

阅读量2.6w

点赞数 17

分类专栏：机器学习文章标签： one-hot编码 one-hot 数组转换one-hot keras

本文链接：https://blog.csdn.net/u011311291/article/details/82763093

版权

机器学习专栏收录该内容

85 篇文章

订阅专栏

import numpy as np
from keras.utils import to_categorical

data = [1, 2, 3, 4, 5, 6, 7, 8, 9, 7]
data = array(data)
print(data)
# [1 2 3 4 5 6 7 8 9 7]

#有普通np数组转换为one-hot
one_hots = to_categorical(data)
print(one_hots)
# [[ 0.  1.  0.  0.  0.  0.  0.  0.  0.  0.]
#  [ 0.  0.  1.  0.  0.  0.  0.  0.  0.  0.]
#  [ 0.  0.  0.  1.  0.  0.  0.  0.  0.  0.]
#  [ 0.  0.  0.  0.  1.  0.  0.  0.  0.  0.]
#  [ 0.  0.  0.  0.  0.  1.  0.  0.  0.  0.]
#  [ 0.  0.  0.  0.  0.  0.  1.  0.  0.  0.]
#  [ 0.  0.  0.  0.  0.  0.  0.  1.  0.  0.]
#  [ 0.  0.  0.  0.  0.  0.  0.  0.  1.  0.]
#  [ 0.  0.  0.  0.  0.  0.  0.  0.  0.  1.]
#  [ 0.  0.  0.  0.  0.  0.  0.  1.  0.  0.]]

# 由one-hot转换为普通np数组
data = [argmax(one_hot)for one_hot in one_hots]
print(data)
# [1, 2, 3, 4, 5, 6, 7, 8, 9, 7]