Tensorflow--取tensorf指定列的操作

In [1]: import os

In [2]: os.environ["CUDA_VISIBLE_DEVICES"] = "0"

In [3]: import tensorflow as tf

In [4]:sess =tf.Session()

In [5]: input = tf.constant([[[1,2,3],[4,5,6],[7,8,9]],[[10,11,12],[13,14,15],[1
   ...: 6,17,18]]])

In [6]: input.get_shape()
Out[6]: TensorShape([Dimension(2), Dimension(3), Dimension(3)])

In [7]: input_2 = input[:,:,2]

In [8]: print(sess.run(input_2))
[[ 3  6  9]
 [12 15 18]]

In [9]: input_2 = input[:,:,0:2]

In [10]: print(sess.run(input_2))
[[[ 1  2]
  [ 4  5]
  [ 7  8]]

 [[10 11]
  [13 14]
  [16 17]]]

In [11]: input = tf.constant([[[[1,2,3],[4,5,6],[7,8,9]],[[10,11,12],[13,14,15],
    ...: [16,17,18]]]])

In [12]: input.get_shape()
Out[12]: TensorShape([Dimension(1), Dimension(2), Dimension(3), Dimension(3)])

In [13]: input_2 = input[:,:,2]

In [14]: print(sess.run(input_2))
[[[ 7  8  9]
  [16 17 18]]]

In [15]: input_2 = input[:,:,:,2]

In [16]: print(sess.run(input_2))
[[[ 3  6  9]
  [12 15 18]]]



 

你可能感兴趣的:(Tensorflow--取tensorf指定列的操作)