Python – tensorflow.dynamic_stitch()

TensorFlow is open-source Python library designed by Google to develop Machine Learning models and deep learning  neural networks.

dynamic_stitch() is used to merge multiple tensors into single tensor.

Syntax: tensorflow.dynamic_stitch( indices, data, name)

Parameter:

• indices: It is a list of Tensors having minimum 1 tensor and each tensor with dtype int32.
• data :  It is list of Tensors having same length as indices.
• name(optional):  It defines the name for the operation.

Result:

It returns  a Tensor of same dtype as data.

Example 1: 

Python3

# Importing the library
import tensorflow as tf

# Initializing the input
indices = [[0, 1, 5], [2, 4, 3, 6]]
data = [[1, 2, 3], [4, 5, 6, 7]]

# Printing the input
print('indices:', indices)
print('data: ', data)

# Calculating result
x = tf.dynamic_stitch(indices, data)

# Printing the result
print('x: ', x)

Output:

indices: [[0, 1, 5], [2, 4, 3, 6]]
data:  [[1, 2, 3], [4, 5, 6, 7]]
x:  tf.Tensor([1 2 4 6 5 3 7], shape=(7, ), dtype=int32)

Example 2:

Python3

# Importing the library
import tensorflow as tf

# Initializing the input
indices = [[0, 1, 6], [5, 4, 3]]
data = [[1, 2, 3], [4, 5, 6]]

# Printing the input
print('indices:', indices)
print('data: ', data)

# Calculating result
x = tf.dynamic_stitch(indices, data)

# Printing the result
print('x: ', x)

Output:

indices: [[0, 1, 2], [5, 4, 3]]
data:  [[1, 2, 3], [4, 5, 6]]
x:  tf.Tensor([1 2 3 6 5 4], shape=(6, ), dtype=int32)