Reverse One-Hot Encode
Published on
Updated on
Warning: This post has not been modified
for over 2 years. For technical posts,
make sure that it is still relevant.
Let’s say that you have a dataset that is one hot encoded like the following observation:
import numpy as np
obs = np.array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0])
The easiest way to reverse one-hot encode the structure, is to take the argmax of the observation.
reverse_encoding = np.argmax(obs)
# 13