https://github.com/masterscrat/getting-started

Introductory notebooks for the Flatland environment and challenge
https://github.com/masterscrat/getting-started

competitive-programming machine-learning operations-research reinforcement-learning

Last synced: 6 months ago
JSON representation

Introductory notebooks for the Flatland environment and challenge

• Host: GitHub
• URL: https://github.com/masterscrat/getting-started
• Owner: MasterScrat
• Created: 2020-01-21T11:01:14.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2020-03-14T04:45:37.000Z (over 5 years ago)
• Last Synced: 2025-03-25T19:15:21.955Z (6 months ago)
• Topics: competitive-programming, machine-learning, operations-research, reinforcement-learning
• Language: Jupyter Notebook
• Homepage: https://www.aicrowd.com/challenges/flatland-challenge
• Size: 1.27 MB
• Stars: 5
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Getting Started with Flatland

> **Flatland** is an environment for developing and comparing multi-agent reinforcement learning algorithms in gridworlds.

This repository contains notebooks to get you started on the right track with the Flatland environment, in order to take part in the [AIcrowd Flatland Challenge](https://www.aicrowd.com/challenges/flatland-challenge).

If you want to dive into challenge baselines right away, [check out the various approaches below](#challenge-baselines). 

Discovering Flatland

---

**Part 1: The Rail Environment**

[![Open In Binder](https://mybinder.org/static/images/badge_logo.svg)](https://mybinder.org/v2/gh/MasterScrat/getting-started/master?filepath=notebook_1.ipynb)

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/MasterScrat/getting-started/blob/master/notebook_1.ipynb)

- Create a `RailEnv` environment and render it

- Check out the default observations

- "Train" a random agent

![notebook1](assets/sparse_env.png)

**Part 2: Observations**

[![Open In Binder](https://mybinder.org/static/images/badge_logo.svg)](https://mybinder.org/v2/gh/MasterScrat/getting-started/master?filepath=notebook_2.ipynb)

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/MasterScrat/getting-started/blob/master/notebook_2.ipynb)

- Finding suitable observations

- Creating your own observations

- Visualizing observations

![notebook2](assets/movie.gif)

**Part 3: Level Generation**

[![Open In Binder](https://mybinder.org/static/images/badge_logo.svg)](https://mybinder.org/v2/gh/MasterScrat/getting-started/master?filepath=notebook_1.ipynb)

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/MasterScrat/getting-started/blob/master/notebook_3.ipynb)

- Creating random rail networks

- Creating schedules

- Adjusting size and difficulty

**Part 4: Malfunctions**

[![Open In Binder](https://mybinder.org/static/images/badge_logo.svg)](https://mybinder.org/v2/gh/MasterScrat/getting-started/master?filepath=notebook_1.ipynb)

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/MasterScrat/getting-started/blob/master/notebook_4.ipynb)

- Introducing stochastic malfunctions

- Handling malfunctions

**Part 5: Speed Profiles**

[![Open In Binder](https://mybinder.org/static/images/badge_logo.svg)](https://mybinder.org/v2/gh/MasterScrat/getting-started/master?filepath=notebook_1.ipynb)

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/MasterScrat/getting-started/blob/master/notebook_5.ipynb)

- Handling agent speed

- Handling partial moves

Challenge Baselines - Coming soon!

---

The Flatland Challenge can be approached in different ways - for example using methods from **operations research**, **reinforcement learning**, or anything else!

The following notebooks show how to approach the problem using each of these methods.

**Reinforcement Learning: DDQN**

[![Open In Binder](https://mybinder.org/static/images/badge_logo.svg)](https://mybinder.org/v2/gh/MasterScrat/getting-started/master?filepath=notebook_1.ipynb)

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/MasterScrat/getting-started/blob/master/notebook_5.ipynb)

Solve simple rail problems using Double DQN.

**Operations Research**

[![Open In Binder](https://mybinder.org/static/images/badge_logo.svg)](https://mybinder.org/v2/gh/MasterScrat/getting-started/master?filepath=notebook_1.ipynb)

[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/MasterScrat/getting-started/blob/master/notebook_5.ipynb)

Solve simple rail problems using OR methods.