python选取tensor某一维_Tensorflow--取tensorf指定列的操作方式

我就废话不多说了，大家还是直接看代码吧~

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),