tf.stack()、tf.unstack()函数介绍和示例

tf.stack()、tf.unstack()函数介绍和示例

1. tf.stack([A, B] axis=0)

释义：矩阵拼接

• A，B，输入张量
• axis，指定拼接维度，默认为0。

示例1：二维数据中，维度 0 处拼接

import tensorflow as tf
 
A = [[1, 2, 3], 
     [4, 5, 6]]
B = [[10, 20, 30], 
     [40, 50, 60]]
X = tf.stack([A, B], axis=0)    # 0维拼接
 
with tf.Session() as sess:
    print(sess.run(X))

[[[ 1  2  3]
  [ 4  5  6]]

 [[10 20 30]
  [40 50 60]]]

示例2：二维数据中，维度 1 处拼接

import tensorflow as tf
 
A = [[1, 2, 3], 
     [4, 5, 6]]
B = [[10, 20, 30], 
     [40, 50, 60]]
X = tf.stack([A, B], axis=1)   # 1维拼接
 
with tf.Session() as sess:
    print(sess.run(X))

[[[ 1  2  3]
  [10 20 30]]

 [[ 4  5  6]
  [40 50 60]]]

示例3：二维数据中，维度 2 处拼接

import tensorflow as tf
 
A = [[1, 2, 3], 
     [4, 5, 6]]
B = [[10, 20, 30], 
     [40, 50, 60]]
X = tf.stack([A, B], axis=2)   # 2维拼接
 
with tf.Session() as sess:
    print(sess.run(X))

[[[ 1 10]
  [ 2 20]
  [ 3 30]]

 [[ 4 40]
  [ 5 50]
  [ 6 60]]]

2. tf.unstack(A, axis=0)

释义：拆分矩阵

• A，输入张量
• axis，指定维度，默认为 0。二维数据中，若为 0，则按行拆分；若为 1，则按列拆分

示例1：二维数据按行拆分

import tensorflow as tf
 
A = [[1, 2, 3], 
     [4, 5, 6]]
X = tf.unstack(A, axis=0)
 
with tf.Session() as sess:
    print(sess.run(X))

[array([1, 2, 3]), array([4, 5, 6])]

示例2：二维数据按列拆分

import tensorflow as tf
 
A = [[1, 2, 3], 
     [4, 5, 6]]
X = tf.unstack(A, axis=1)
 
with tf.Session() as sess:
    print(sess.run(X))

[array([1, 4]), array([2, 5]), array([3, 6])]