Gym Trading Env is a Gymnasium environment for simulating stocks and training Reinforcement Learning (RL) trading agents. It was designed to be fast and customizable for easy RL trading algorithms implementation.

Github

Key features

This package aims to greatly simplify the research phase by offering :

• Easy and quick download technical data on several exchanges

• A simple and fast environment for the user and the AI, but which allows complex operations (Short, Margin trading).

• A high performance rendering (can display several hundred thousand candles simultaneously), customizable to visualize the actions of its agent and its results.

• (Coming soon) An easy way to backtest any RL-Agents or any kind

Installation

Gym Trading Env supports Python 3.9+ on Windows, Mac, and Linux. You can install it using pip:

pip install gym-trading-env

Or using git :

git clone https://github.com/ClementPerroud/Gym-Trading-Env

Contents

• Introduction
• Gettings Started
  • Installation
  • Import
• Environment Quick Summary
  • Important Parameters
  • Action Space
  • Observation Space
  • Reward
  • Starting State
  • Episode Termination
  • Arguments

🤖 Reinforcement Learning

• Tutorial
  • Understand the action space
    • Positions
    • Complex positions
  • Market data
    • Import your own dataset
    • Easy download for crypto
  • Create your features
  • Create your first environment
  • Run the environment
• Customization
  • Custom reward function
  • Custom logs
• Features
  • Create static features
  • Create dynamic features
• Multi datasets environment
  • Multi Dataset Trading Environment
    • How to use ?
    • Easy preprocess
    • Run the environment
• Vectorize your env
  • Gym vector
  • Special cases
  • Run the environments

🦾 Functionnalities

• Render
  • Your first render
  • Customize your render
    • Add custom lines
    • Add custom metrics
• Download market data
  • Crypto market data
    • Supported exchanges …
    • More exchanges …
  • Stock market data

📚 Reference

• History Object
• Environments