tf.reduce_sum 语法

tf.reduce_sum：


  1 #coding=utf8
  2 import tensorflow as tf
  3 import numpy as np
  4
  5 tf.compat.v1.disable_eager_execution()
  6
  7 x = np.array([[[1, 2, 3], [4, 5, 6]], [[7, 8, 9], [10, 11, 12]]]) # shape = 2*2*3
  8 x_p = tf.compat.v1.placeholder(dtype=tf.int32, shape=[None, None, None], name='input')
  9
 10 #y =  tf.reduce_sum(x_p, 0) # 按行
 11 #y =  tf.reduce_sum(x_p, 1) #按列
 12 y =  tf.reduce_sum(x_p, 2) #修改这里
 13
 14
 15 with tf.compat.v1.Session() as sess:
 16     y = sess.run(y, feed_dict={x_p: x})
 17     print(y)

tf.reduce_sum(x_p, 0):

[[ 8 10 12]
 [14 16 18]]

y =  tf.reduce_sum(x_p, 1)

[[ 5  7  9]
 [17 19 21]]

tf.reduce_sum(x_p, 2)

[[ 6 15]
 [24 33]]

 

tf.reduce_mean：

import tensorflow as tf
import numpy as np
 
x = tf.constant([[1,2],[3,4]])
with tf.Session() as sess:
    print(sess.run(tf.reduce_mean(x))) #所有求平均
    print(sess.run(tf.reduce_mean(x, 0))) #按 行 求和
    print(sess.run( tf.reduce_mean(x, 1)))# 按 列 求平均
    print(sess.run(tf.reduce_max(x)))
    print(sess.run(tf.reduce_max(x, 0)))
    print(sess.run(tf.reduce_max(x, 1)))

#########输出##########
2
[2 3]
[1 3]
4
[3 4]
[2 4]

 

[tf]tf.gather_nd的用法

https://www.jianshu.com/p/6e4398d9b4c8

 

tf.transpose

tf.transpose的第二个参数perm=[0,1,2],0代表三维数组的高（即为二维数组的个数），1代表二维数组的行，2代表二维数组的列。
tf.transpose(x, perm=[1,0,2])代表将三位数组的高和行进行转置