tf.reduce_mean的用法

import tensorflow as tf;
import numpy as np;

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

with tf.Session() as sess:
	print sess.run(tf.reduce_mean(A))
	print sess.run(tf.reduce_mean(A, axis=0))
	print sess.run(tf.reduce_mean(A, axis=1))

输出：

2 #整体的平均值
[2 3] #按列求得平均
[1 3] #按照行求得平均

如果没有axis求得就是整体的平均，