tensorflow.tile() 函数

tile( input, multiples, name=None )     按multiples设定的各维度扩展倍数进行数据扩展

参数：input --> 输入tensor

            multiples   --> 指定各维度要扩展的倍数

            name -->名称（可选）

import tensorflow as tf

data = tf.Variable(tf.random_normal(shape=(1, 2, 3)))
tile = tf.tile(data, multiples=[3, 2, 2])

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    print('原始数据：')
    print(sess.run(data))
    print('扩展后的数据：')
    print(sess.run(tile))

原始数据：
[[[ 0.08998119 -0.45589316  0.30437934]
  [ 0.80043739 -0.34145349 -0.34891185]]]
扩展后的数据：
[[[ 0.08998119 -0.45589316  0.30437934  0.08998119 -0.45589316  0.30437934]
  [ 0.80043739 -0.34145349 -0.34891185  0.80043739 -0.34145349 -0.34891185]
  [ 0.08998119 -0.45589316  0.30437934  0.08998119 -0.45589316  0.30437934]
  [ 0.80043739 -0.34145349 -0.34891185  0.80043739 -0.34145349 -0.34891185]]

 [[ 0.08998119 -0.45589316  0.30437934  0.08998119 -0.45589316  0.30437934]
  [ 0.80043739 -0.34145349 -0.34891185  0.80043739 -0.34145349 -0.34891185]
  [ 0.08998119 -0.45589316  0.30437934  0.08998119 -0.45589316  0.30437934]
  [ 0.80043739 -0.34145349 -0.34891185  0.80043739 -0.34145349 -0.34891185]]

 [[ 0.08998119 -0.45589316  0.30437934  0.08998119 -0.45589316  0.30437934]
  [ 0.80043739 -0.34145349 -0.34891185  0.80043739 -0.34145349 -0.34891185]
  [ 0.08998119 -0.45589316  0.30437934  0.08998119 -0.45589316  0.30437934]
  [ 0.80043739 -0.34145349 -0.34891185  0.80043739 -0.34145349 -0.34891185]]]