tf.reshape()

tf.reshape

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

例如:

  • 将 1x9矩阵 ==> 3x3矩阵
import tensorflow as tf
import numpy as np

A = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9])

t = tf.reshape(A, [3, 3])
with tf.Session() as sess:
    print(sess.run(t))

# 输出
[[1, 2, 3],
[4, 5, 6],
[7, 8, 9]]
  • 将 3x2x3矩阵 ==> 1x18矩阵(也就是整个矩阵平铺)
import tensorflow as tf
import numpy as np

A = np.array([[[1, 1, 1],
               [2, 2, 2]],
              [[3, 3, 3],
               [4, 4, 4]],
              [[5, 5, 5],
               [6, 6, 6]]])

t = tf.reshape(A, [-1])
with tf.Session() as sess:
    print(sess.run(t))

# 输出
[1 1 1 2 2 2 3 3 3 4 4 4 5 5 5 6 6 6]
  • 将 3x2x3矩阵 ==> 2x9矩阵([2, -1]表示列项平铺)
import tensorflow as tf
import numpy as np

A = np.array([[[1, 1, 1],
               [2, 2, 2]],
              [[3, 3, 3],
               [4, 4, 4]],
              [[5, 5, 5],
               [6, 6, 6]]])

t = tf.reshape(A, [2, -1])
with tf.Session() as sess:
    print(sess.run(t))

# 输出
[[1 1 1 2 2 2 3 3 3]
 [4 4 4 5 5 5 6 6 6]]
  • 将 3x2x3矩阵 ==> 2x3x3矩阵([2, -1, 3]表示行项平铺)
import tensorflow as tf
import numpy as np

A = np.array([[[1, 1, 1],
               [2, 2, 2]],
              [[3, 3, 3],
               [4, 4, 4]],
              [[5, 5, 5],
               [6, 6, 6]]])

t = tf.reshape(A, [2, -1, 3])
with tf.Session() as sess:
    print(sess.run(t))

# 输出
[[[1 1 1]
  [2 2 2]
  [3 3 3]]

 [[4 4 4]
  [5 5 5]
  [6 6 6]]]
  • 将1x1矩阵(只能为1x1的矩阵,否则形状不符,Occur ValueError) ==> Scalar(标量),也就是一个数
import tensorflow as tf
import numpy as np

A = np.array([7])

t = tf.reshape(A, [])
with tf.Session() as sess:
    print(sess.run(t))

# 输出
7

参数

  • tensor:输入的张量
  • shape:表示重新设置的张量形状,必须是int32或int64类型
  • name:表示这个op名字,在tensorboard中才会用

返回值

  • 有着重新设置过形状的张量
©️2020 CSDN 皮肤主题: 编程工作室 设计师:CSDN官方博客 返回首页