tearitco/deep-rts

A Real-Time-Strategy game for Deep Learning research

Description

A Real-Time Strategy Simulator for (Deep) Reinforment Learning in Python and C++! It is recommended to use the master-branch for the newest (and usually best) version of the environment. I am greatful for any input in regards to improving the environment.

Please use the following citation when using this in your work!

@INPROCEEDINGS{8490409,
author={P. {Andersen} and M. {Goodwin} and O. {Granmo}},
booktitle={2018 IEEE Conference on Computational Intelligence and Games (CIG)},
title={Deep RTS: A Game Environment for Deep Reinforcement Learning in Real-Time Strategy Games},
year={2018},
volume={},
number={},
pages={1-8},
keywords={computer games;convolution;feedforward neural nets;learning (artificial intelligence);multi-agent systems;high-performance RTS game;artificial intelligence research;deep reinforcement learning;real-time strategy games;computer games;RTS AIs;Deep RTS game environment;StarCraft II;Deep Q-Network agent;cutting-edge artificial intelligence algorithms;Games;Learning (artificial intelligence);Machine learning;Planning;Ground penetrating radar;Geophysical measurement techniques;real-time strategy game;deep reinforcement learning;deep q-learning},
doi={10.1109/CIG.2018.8490409},
ISSN={2325-4270},
month={Aug},}

Dependencies

• Python >= 3.9.1

Installation

Method 1 (From Git Repo)

sudo pip3 install git+https://github.com/cair/DeepRTS.git

Method 2 (Clone & Build)

git clone https://github.com/cair/deep-rts.git
cd deep-rts
git submodule sync
git submodule update --init
sudo pip3 install .

Available maps

10x10-2-FFA
15x15-2-FFA
21x21-2-FFA
31x31-2-FFA
31x31-4-FFA
31x31-6-FFA

Minimal Example

import random
from DeepRTS import python
from DeepRTS import Engine

from DeepRTS.python import scenario

if __name__ == "__main__":

    episodes = 10000000
    random_play = True
    gui_config = python.Config(
        render=True,
        view=True,
        inputs=True,
        caption=False,
        unit_health=True,
        unit_outline=False,
        unit_animation=True,
        audio=True,
        audio_volume=50
    )

    engine_config: Engine.Config = Engine.Config.defaults()
    engine_config.set_barracks(True)
    engine_config.set_footman(True)
    engine_config.set_instant_town_hall(True)
    engine_config.set_archer(True)
    engine_config.set_start_wood(2000)
    engine_config.set_start_gold(2000)
    engine_config.set_start_oil(2000)

    game = python.Game(
        python.Config.Map.FIFTEEN,
        n_players=1,  # TODO - Only 1 for some scenarios?
        engine_config=engine_config,
        gui_config=gui_config,
        terminal_signal=False
    )
    game.set_max_fps(30)
    game.set_max_ups(10000000)

    env = scenario.GoldThousand(game)

    for episode in range(episodes):
        print("Episode: %s, FPS: %s, UPS: %s" % (episode, game.get_fps(), game.get_ups()))

        terminal = False
        state = env.reset()
        while not terminal:
            action = random.randint(0, 15)  # TODO AI Goes here
            next_state, reward, terminal, _ = env.step(action, render="human")

            state = next_state

In-Game Footage

10x10 - 2 Player - free-for-all

15x15 - 2 Player - free-for-all

21x21 - 2 Player - free-for-all

31x31 - 2 Player - free-for-all

31x31 - 4 Player - free-for-all

31x3 - 6 Player - free-for-all