Reverse One-Hot Encode

Brandon Rozek

October 11, 2020

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