{"url":"/dataset/maclight","name":"MacLight","full_name":"MacLight_4×4_grid","description_markdown":"If you plan to test your method on our road network, you can find road network files in `env\\map`.\r\n\r\n- `ff.net.xml` is road network. It is unchanging, and the difference between the different environments is the traffic flow file.\r\n- `ff_normal.rou.xml` is traffic flow file under normal pressure (Normal&Block).\r\n- `ff_hard.rou.xml` is traffic flow file under high pressure (Peak).\r\n\r\nIt should be noted that the normal pressure we set is also relatively high.\r\n\r\n\r\n|          | Normal&Block | Peak | arterial4x4 | grid4x4 |\r\n|----------|--------------|------|-------------|---------|\r\n| Vehicles | 8000         | 10286| 2485        | 1472    |\r\n\r\n\r\nExamples to create an enviroment\r\n\r\nOur environment interface inherits from [gymnasium](https://gymnasium.farama.org/) and [sumo-rl](https://github.com/LucasAlegre/sumo-rl), so you can easily migrate your algorithms.\r\n\r\n1. Static environment:\r\n\r\n   The direction of the vehicle is completely fixed and the route will not be changed.\r\n\r\n```python\r\nenv = sumo_rl.parallel_env(\r\n    net_file='env/map/ff.net.xml',\r\n    route_file=f'env/map/ff_normal.rou.xml',  # Could be `ff_hard.rou.xml`\r\n    num_seconds=args.seconds,\r\n    use_gui=False,\r\n    sumo_warnings=False,\r\n    additional_sumo_cmd='--no-step-log'\r\n)\r\n```\r\n\r\n2. Dynamic environment\r\n\r\n   We define the `BlockStreet` class, which can randomly block certain roads, so that vehicles reselect the best route, which will cause sudden changes in traffic flow on certain roads.\r\n\r\n```python\r\nfrom env.wrap.random_block import BlockStreet\r\n# block_num: Number of blocked roads, like 8\r\n# seconds: Simulation seconds, up to 3600\r\nenv = BlockStreet(env, block_num, seconds)\r\n```\r\n\r\nAdditionally, we strongly recommend that you set the following environment variables to get the fastest possible simulation (although it may still be slower):\r\n\r\n```python\r\nimport os\r\nos.environ['LIBSUMO_AS_TRACI'] = '1'\r\n```","description_withheld":null,"homepage":"https://github.com/Aegis1863/MacLight","introduced_date":"2024-12-20","introduced_date_note":null,"introduced_by":{"paper":"/paper/maclight-multi-scene-aggregation","title":"MacLight: Multi-scene Aggregation Convolutional Learning for Traffic Signal Control","first_author":"Sunbowen Lee","url":null},"license":null,"modalities":[],"tasks":[],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["MacLight"],"data_loaders":[],"num_papers_in_archive":1,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[],"papers_with_a_benchmark_row":[],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":0,"samples_harvested":0,"samples_ran":0,"samples_unverified":0,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}