website/content/blog/reverseonehotencode.md

23 lines
480 B
Markdown
Raw Normal View History

2020-10-12 02:04:08 +00:00
---
title: "Reverse One-Hot Encode"
date: 2020-10-11T21:58:47-04:00
draft: false
2022-01-02 19:24:29 +00:00
tags: ["Python"]
2020-10-12 02:04:08 +00:00
---
Let's say that you have a dataset that is one hot encoded like the following observation:
```python
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.
```python
reverse_encoding = np.argmax(obs)
# 13
```