SUMO RL Project Setup

This guide provides step-by-step instructions to set up the SUMO RL environment, create a Conda environment, install the required Python modules, and run experiments. It also includes details on generating and analyzing plots from experiment outputs.

---

Prerequisites

Before starting, ensure you have the following installed on your system:

• Python (version 3.8 or later recommended)
• Anaconda/Miniconda
• SUMO (Simulation of Urban MObility) installed and added to your system's PATH

---

Step 1: Clone the Repository

First, clone the repository to your local machine:

git clone https://github.com/Transyltooniaa/RL-Project.git
cd RL-Project

---

Step 2: Create a Conda Environment

1. Create a new Conda environment for the project:

   conda create -n sumo_rl_env python=3.8 -y

2. Activate the environment:

   conda activate sumo_rl_env

---

Step 3: Install Necessary Python Modules

Install the required Python modules listed in the requirements.txt file:

pip install -r requirements.txt

For PettingZoo, install a specific version:

pip uninstall pettingzoo -y
pip install pettingzoo==1.23.1

To ensure all dependencies are installed, you can also run:

pip install argparse pandas numpy tqdm gymnasium traci supersuit pyvirtualdisplay sumo_rl stable-baselines3 ray[rllib] tensorflow-probability torch

---

Step 4: Setting Up SUMO RL Environment

Ensure that SUMO is correctly installed and accessible from your terminal. You can verify this by running the following command:

sumo --version

If SUMO is not installed or not in your PATH, refer to the SUMO documentation for installation guidance.

Configure SUMO_HOME and SUMO_PATH on Windows

1. Locate the installation directory for SUMO, typically:

   C:\Program Files (x86)\Eclipse\Sumo
   

2. Set the environment variables:
   • SUMO_HOME: Point to the main directory, e.g., C:\Program Files (x86)\Eclipse\Sumo
   • SUMO_PATH: Add the bin sub-directory to your system's PATH, e.g., C:\Program Files (x86)\Eclipse\Sumo\bin
3. Steps to configure:
   • Open System Properties > Advanced > Environment Variables.
   • Under System Variables, select New.
   • Add:
     • Variable name: SUMO_HOME
     • Variable value: C:\Program Files (x86)\Eclipse\Sumo
   • Find the Path variable, click Edit, and add: C:\Program Files (x86)\Eclipse\Sumo\bin.

---

Step 5: Running Experiments

To run an experiment, navigate to the experiment directory and use the following command:

python experiment/<file_name>

Replace <file_name> with the name of the Python script you want to execute.

Example:

python experiment/train_agent.py

The experiment outputs are saved as CSV files in the specified outputs directory within the repository.

---

Step 6: Generating and Analyzing Plots

After running experiments, you can visualize and analyze the results using the plot.py script.

About plot.py

• This script generates plots based on the CSV outputs generated by experiments.
• It looks for CSV files stored in the outputs folder, which is created automatically after you run experiments.
• The script allows you to compare one column of a CSV file with another, providing a clear visualization of the data.

Generate a Plot

To generate a plot, run the following command:

python outputs/plot.py

Example Usage

• Use this script to analyze and compare metrics such as rewards, waiting times, or other key performance indicators stored in the CSV files.
• Visualizations can be saved for later use.

---

Additional Dependencies

In addition to the dependencies listed in requirements.txt, the following modules are used across the project:

• argparse
• os, sys, shutil, subprocess
• pandas
• numpy
• tqdm
• gymnasium
• traci
• supersuit
• pyvirtualdisplay.smartdisplay
• sumo_rl
• stable_baselines3
• ray[rllib]
• tensorflow-probability
• torch

Ensure all these dependencies are installed for the project to run smoothly.

---

Additional Notes

• If you encounter any issues, ensure that all required modules are installed and that SUMO is properly set up.