{"id":21820793,"url":"https://github.com/mryellow/maze_explorer","last_synced_at":"2025-04-14T03:01:28.256Z","repository":{"id":57440195,"uuid":"97892485","full_name":"mryellow/maze_explorer","owner":"mryellow","description":"A simple maze exploration game for AI agents","archived":false,"fork":false,"pushed_at":"2019-09-29T00:21:39.000Z","size":446,"stargazers_count":6,"open_issues_count":1,"forks_count":4,"subscribers_count":3,"default_branch":"master","last_synced_at":"2024-04-29T23:48:13.475Z","etag":null,"topics":["game","maze","maze-explorer","openai-gym","openai-gym-environments"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/mryellow.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2017-07-21T01:25:31.000Z","updated_at":"2023-03-04T15:39:24.000Z","dependencies_parsed_at":"2022-09-26T17:11:35.386Z","dependency_job_id":null,"html_url":"https://github.com/mryellow/maze_explorer","commit_stats":null,"previous_names":[],"tags_count":4,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mryellow%2Fmaze_explorer","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mryellow%2Fmaze_explorer/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mryellow%2Fmaze_explorer/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/mryellow%2Fmaze_explorer/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/mryellow","download_url":"https://codeload.github.com/mryellow/maze_explorer/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248813789,"owners_count":21165633,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["game","maze","maze-explorer","openai-gym","openai-gym-environments"],"created_at":"2024-11-27T16:39:14.086Z","updated_at":"2025-04-14T03:01:28.219Z","avatar_url":"https://github.com/mryellow.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Maze Explorer\n\nA maze exploration game for AI agents.\n\n![Screenshot](https://raw.githubusercontent.com/mryellow/maze_explorer/master/mazeexp/engine/assets/screen_001.jpg)\n\n## TODO:\n\n* [ ] New game modes\n* [ ] Terminal state for human, with switch for god-mode\n* [ ] Zoomed scrolling view for human players\n\n## Installation\n\n### Source\n\n```bash\ngit clone https://github.com/mryellow/maze_explorer.git\ncd maze_explorer\npip install -e .\n```\n\n### Package\n\n```bash\npip install mazeexp\n```\n\n## Standalone\n\n`python standalone.py`\n\n### Options\n\n```bash\n--mode X # Mode number\n--random # Execute random actions step-by-step via `act`\n--step # Call the engine step-by-step via `step`\n```\n\n\n## OpenAIGym\n\n[gym-mazeexplorer](https://github.com/mryellow/gym-mazeexplorer)\n\n### Game modes\n\n#### Mode 0 `MazeExplorerEat-v0`\n\nApples and poison.\n\nBased on [Andrej Karpathy's Javascript environment](https://cs.stanford.edu/people/karpathy/convnetjs/demo/rldemo.html)\n\n##### State\n\n```\n[\n  [min(wall_range, apple_range, poison_range), apple_range, poison_range],\n  ...\n]\n```\n\n##### Rewards\n\n* ~~`avg(proximity) or 1` Agents don't like seeing walls, especially up close~~\n* ~~`*1.1` Forward action bonus~~\n* `-10` collision with wall\n* `+5` collision with apple\n* `-6` collision with poison\n\n##### Terminal\n\n* None\n\n#### Mode 1 `MazeExplorerExplore-v0`\n\nExplore the maze and make it back to spawn before battery runs out.\n\n##### State\n\n```\n[wall_range, ..., battery]\n```\n\n##### Rewards\n\n* `-100` collision with wall\n* `-100` battery out\n\n###### When battery above 50%\n\n* `+1` exploration reward on first visit to tile and for each of it's open neighbours\n\nThe `visited` state is not observable in environment and reward is generated by ground-truth.\nThus agent must keep it's own internal state and/or develop a policy which overcomes this unknown.\n\n###### When battery below 50%\n\n* `+200` reward goal state on reaching spawn tile\n* No futher exploration rewards\n\nSpawn tile is no different to any other from agents perspective, must remember how to return to it or develop a policy which increases the likelihood of such.\n\n##### Terminal\n\n* Wall collision\n* Battery out\n* Return to home goal\n\n## Cite\n\nIf you use Maze Explorer in your academic research, we would be grateful if you could cite it as follows:\n\n```\n@misc{mryellow2017mazeexplorer,\n    author = {Mr-Yellow},\n    title = {Maze Explorer: A maze exploration game for AI agents},\n    howpublished={Web page},\n    url = {https://github.com/mryellow/maze_explorer},\n    year = {2017}\n}\n```\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmryellow%2Fmaze_explorer","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmryellow%2Fmaze_explorer","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmryellow%2Fmaze_explorer/lists"}