tensorflow tenosr的操作 concat 和 reshape

1、concat的操作是将矩阵逐行拼起来,或者逐列拼起来：

import tensorflow as tf

t1 = [[1,2,3], [4,5,6]]
t2 = [[7,8,9], [10,11,12]]
sess = tf.Session()
print (sess.run(tf.concat(1, [t1, t2])))

结果：concat(1,[t1,t2])进行逐行拼凑,在LSTM中很经常用,每个时刻的输入都要产生输出,然后我们要将输出逐行拼起来,形成整个num_steps个时刻的输出
[[ 1  2  3  7  8  9]
 [ 4  5  6 10 11 12]]

import tensorflow as tf

t1 = [[1,2,3], [4,5,6]]
t2 = [[7,8,9], [10,11,12]]
sess = tf.Session()
print (sess.run(tf.concat(0, [t1, t2])))

结果：tf.concat(0, [t1, t2])进行逐列输出

[[ 1  2  3]
 [ 4  5  6]
 [ 7  8  9]
 [10 11 12]]

===============================================================================================================================

2、reshape的操作

import tensorflow as tf

tensor = tf.constant([1, 2, 3, 4, 5, 6, 7,8])
sess = tf.Session()
sess.run(tf.initialize_all_variables())
print('--------------')
print(sess.run(tensor))

tensorReshape = tf.reshape(tensor,[2,4])
print('---------------------')
print(sess.run(tensorReshape))

tensorReshape = tf.reshape(tensor,[-1,2,2])
print('---------------------')
print(sess.run(tensorReshape))

结果：

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


 [[5 6]
  [7 8]]]

说明:

所以-1，就是缺省值，就是先以你们合适，到时总数除以你们几个的乘积，我该是几就是几

===============================================================================================================================