• Changelog

• Documentation

A plugin for Python SC2 API BurnySc2

This module is inspired by plays like this one TY map positioning (notice how the army splits into groups, covering different areas, tanks are tucked in corners, and so on)

Hopefully with the interface provided here, you will be able to build plays like that one!

Thanks A lot to DrInfy for solving one of the biggest challenges, finding rare choke points.

Check out his work

• Sharpy for rapid bot development.

• sc2pathlib a high performance rust module with python interface for pathfinding

More Examples reside in the Documentation

See here for an example map_analyzer reaper bot showing pathing and influence in action.

Example:

import pickle
import lzma
from map_analyzer import MapData
from map_analyzer.utils import import_bot_instance

#if its from BurnySc2 it is compressed
# https://github.com/BurnySc2/python-sc2/tree/develop/test/pickle_data
YOUR_FILE_PATH = 'some_directory/map_file'
with lzma.open(YOUR_FILE_PATH, "rb") as f:
    raw_game_data, raw_game_info, raw_observation = pickle.load(f)

# mocking a bot object to initalize the map,  this is for when you want to do this while not in a game,  
# if you want to use it in a game just pass in the bot object like shown below 

bot = import_bot_instance(raw_game_data, raw_game_info, raw_observation)


# And then you can instantiate a MapData Object like so
map_data = MapData(bot)


# plot the entire labeled map
map_data.plot_map()

# red dots or X are vision blockers,
# ramps are marked with white dots 
# ramp top center is marked with '^'
# gas geysers are yellow spades 
# MDRampss are marked with R<region_label>
# height span is with respect to :   light = high , dark = low
# ChokeArea is marked with green heart suites
# Corners are marked with a red 'V' 

Tested Maps ( AiArena ) : See map_analyzer/pickle_game_info for all tested maps.

If you already have a BurnySc2 development environment setup, you're likely equipped with all the necessary dependencies Therefore, integrating map_analyzer into your existing bot is a straightforward process requiring just a few simple steps.

If you're a new bot author, please set up a new BurnySc2 bot development environment before installing map_analyzer.

1. Clone or download this repo
2. Copy the map_analyzer directory and place it in the root of your bot directory:

MyBot
├── map_analyzer
│ ├── … dependencies for map_analyzer
│ … your other bot files and folders here

3. map_analyzer relies on a pathing extension written in C, this can be built locally or downloaded from github actions. If you're on a debian based OS you may be able to skip this step as the repo contains a linux binary already included in the map_analyzer folder.
   
   Method 1: Without needing C++ build tools
   
   Check the most recent BuildCExtension Github Action workflow. Then scroll to the bottom to download the artifact for your OS:

4. In your bot initiate the map_analyzer module:

from map_analyzer import MapData
from sc2.bot_ai import BotAI

class MyBot(BotAI):
    map_data: MapData

    async def on_start(self) -> None:
        self.map_data = MapData(self)

    async def on_step(self, iteration: int):
        pass

5. Uploading to AiArena and tournaments -
   No further setup is required, just include the map_analyzer folder in your bot zip.

If you're interested in contributing or would like to run tests then the full dev environment should be setup:

1. Install Poetry for example: pip install poetry
2. poetry install --with dev - This will install all development dependencies, build the C extension and create a new environment. Useful poetry environment commands:
   poetry env list --full-path - - Use this to configure your IDE to recognise the environment
   poetry env remove <env_name>
poetry shell
3. Check your environment is working by running the example bot:
   poetry run examples/MassReaper/run.py

(github workflow to check this on PR)

poetry run pytest

(github workflow to check these on PR)

poetry run black .

poetry run isort .

poetry run flake8 .

To faciliatate automated releases, conventional commits guideline should be followed. Example git commits:

Feature: feat: find path with multiple grids

Bugfix: fix: correct weight cost on cliff

Pull request titles should follow these guidelines too, this enables the automatic release and changelogs to work. There is a github workflow to enforce this.

Example PR title:

feat: find path with multiple grids