Python – tensorflow.argsort() method

TensorFlow is open-source python library designed by Google to develop Machine Learning models  and deep learning  neural networks. Tensorflow have a method argsort() which is used to find the indices of a Tensor in sorted order.

Syntax: tf.argsort(values, axis, direction, stable, name)

Parameters:

  • values: It is a numeric Tensor of any dimension.
  • axis: It defines the axis along which shorting need to be done. If no value is given, default is -1 and sorting is done based on last axis.
  • direction: Either ASCENDING or DESCENDING.
  • stable: Either  True or False. If it’s true then in case of equality original order is maintained.
  • name: It’s an optional argument which define the name for the operation.

Return: It returns a Tensor of type int32 having same shape as values.This tensor contains the indices that will give the sorted order of given values.

If axis or direction is invalid it will raise ValueError.

Example 1:



Python3

filter_none

edit
close

play_arrow

link
brightness_4
code

# importing the library
import tensorflow
  
# initializing value
a= [1,5,2.5,10,7,8.5]
  
# getting the indices for sorted values
b = tensorflow.argsort(a,axis=-1,direction='ASCENDING',stable=False)
  
# printing the result
print('Indices:'b)
  
print('Sorted values')
#printing the sorted value
for i in b:
  print(a[i])

chevron_right


Output:

Indices: tf.Tensor([0 2 1 4 5 3], shape=(6,), dtype=int32)
Sorted Values
1
2.5
5
7
8.5
10

Example 2: In this example wrong value is passes to direction. This will raise ValueError.

Python3

filter_none

edit
close

play_arrow

link
brightness_4
code

# importing the library
import tensorflow
  
# initializing value
a= [1,5,2.5,10,7,8.5]
  
# getting the indices for sorted values
b = tensorflow.argsort(a,axis=-1,direction='ASC',stable=False)

chevron_right


Output:

ValueError: ASC should be one of ASCENDING, DESCENDING




My Personal Notes arrow_drop_up

Check out this Author's contributed articles.

If you like GeeksforGeeks and would like to contribute, you can also write an article using contribute.geeksforgeeks.org or mail your article to contribute@geeksforgeeks.org. See your article appearing on the GeeksforGeeks main page and help other Geeks.

Please Improve this article if you find anything incorrect by clicking on the "Improve Article" button below.


Article Tags :

Be the First to upvote.


Please write to us at contribute@geeksforgeeks.org to report any issue with the above content.