[深度学习]零基础入门TensorFlow之reshape

导语

在深度学习中，数据变形非常普遍，reshape作用是将tensor变换为参数shape的形式

语法

tf.reshape(tensor, shape, name=None) 

tensor为入参，shape为变换的矩阵格式，name可略

实例1

import tensorflow as tf

#定义常量，一维矩阵
tensor = tf.constant([1, 2, 3, 4, 5, 6, 7,8])  

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer()) #初始化Tensor变量
    print(sess.run(tensor))  #输出常量
    #转换为两行四列的矩阵，shape=[2,4]
    tensorReshape = tf.reshape(tensor,[2,4])
    print(sess.run(tensorReshape))  #输出变形后

第一个print原矩阵

[1 2 3 4 5 6 7 8]

第二个print输出变换的矩阵

[[1 2 3 4]
 [5 6 7 8]]

shape的列表数决定了转换后的矩阵维度，比如

shape=[2,4]#二维矩阵
shape=[1,2,4]#三维矩阵

shape的列表值可以为-1

实例2

import tensorflow as tf

tensor = tf.constant([1, 2, 3, 4, 5, 6, 7,8])

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    tensorReshape = tf.reshape(tensor,[2,-1])
    print(sess.run(tensorReshape))

输出

[[1 2 3 4]
 [5 6 7 8]]

实例3

import tensorflow as tf

tensor = tf.constant([1, 2, 3, 4, 5, 6, 7,8])

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    tensorReshape = tf.reshape(tensor,[-1,4])
    print(sess.run(tensorReshape))

输出

[[1 2 3 4]
 [5 6 7 8]]

可以看出shape的三种入参形式[2,4][2,-1][-1,4]，得到相同的变换矩阵，
-1代表的含义是不用我们自己指定这一维的大小，函数会自动计算，但列表中只能存在一个-1

附：官方例子

# tensor 't' is [1, 2, 3, 4, 5, 6, 7, 8, 9]
# tensor 't' has shape [9]
reshape(t, [3, 3]) ==> [[1, 2, 3],
                        [4, 5, 6],
                        [7, 8, 9]]

# tensor 't' is [[[1, 1], [2, 2]],
#                [[3, 3], [4, 4]]]
# tensor 't' has shape [2, 2, 2]
reshape(t, [2, 4]) ==> [[1, 1, 2, 2],
                        [3, 3, 4, 4]]

# tensor 't' is [[[1, 1, 1],
#                 [2, 2, 2]],
#                [[3, 3, 3],
#                 [4, 4, 4]],
#                [[5, 5, 5],
#                 [6, 6, 6]]]
# tensor 't' has shape [3, 2, 3]
# pass '[-1]' to flatten 't'
reshape(t, [-1]) ==> [1, 1, 1, 2, 2, 2, 3, 3, 3, 4, 4, 4, 5, 5, 5, 6, 6, 6]

# -1 can also be used to infer the shape

# -1 is inferred to be 9:
reshape(t, [2, -1]) ==> [[1, 1, 1, 2, 2, 2, 3, 3, 3],
                         [4, 4, 4, 5, 5, 5, 6, 6, 6]]
# -1 is inferred to be 2:
reshape(t, [-1, 9]) ==> [[1, 1, 1, 2, 2, 2, 3, 3, 3],
                         [4, 4, 4, 5, 5, 5, 6, 6, 6]]
# -1 is inferred to be 3:
reshape(t, [ 2, -1, 3]) ==> [[[1, 1, 1],
                              [2, 2, 2],
                              [3, 3, 3]],
                             [[4, 4, 4],
                              [5, 5, 5],
                              [6, 6, 6]]]

# tensor 't' is [7]
# shape `[]` reshapes to a scalar
reshape(t, []) ==> 7