CODEWITHSUNDEEP
pythontensorflowone-hot-encoding
AzureIP
AzureIP

Reputation: 395

Why does tensorflow doubles the one_hot on values?

I have the following test program in Python:

import tensorflow as tf
sess = tf.InteractiveSession()
# Some tensor we want to print the value of
a = tf.one_hot(1,5)

# Add print operation
a = tf.Print(a, [a], message = "This is a: ")

# Add more elements of the graph using a
b = tf.add(a, a)

b.eval()

I call the function that shall create a nice one hot encoding. I would expect that the output is:

array([0., 1., 0., 0., 0.], dtype=float32)

But instead the output is:

array([0., 2., 0., 0., 0.], dtype=float32)

Why?

Upvotes: 1

Views: 50

Answers (1)

David Culbreth
David Culbreth

Reputation: 2796

You are adding a to itself, and then printing the addition. So essentially... a= 1; print (a+a) Obviously, that's not how it's written, but I sure hope that 1+1 is 2.

Upvotes: 2

Related Questions