Search code examples
pythontensorflowone-hot-encoding

Tensorflow Initialize list with 0s and 1s similar to tf.one_hot

I currently have a list of values over a time sequence say the values are [1, 3, 5, 7, 3]. Currently I am using tf.one_hot to get a one hot vector/tensor representative for each value within the list.

1 = [0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0]
3 = [0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0]
5 = [0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0]
7 = [0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0]
3 = [0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0]

In Tensorflow is there a way/function that would allow me to do something similar but initialize all values from 0 to the value with 1s?

Desired Result:

1 = [1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0]
3 = [1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
5 = [1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0]
7 = [1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0]
3 = [1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0]
Solution

  • You are looking for tf.sequence_mask.

    TensorFlow API