Skip to content
GitLab
Switch branch/tag
History Find file
Clone
Merge branch 'anagy-master-patch-94006' into 'master'
Attila authored 
Update README.md

See merge request !1
ee9d4ef4
Name Last commit Last update
LICENSE
README.md
README.md

MEDIANETS - Traffic Incident Dataset

To create the incident dataset, the PeMS dataset was used. The Traffic Incident Dataset contains 5-minute aggregation traffic data from District 3 (Sacramento area) for a one year period from January 1, 2016 to December 31, 2016. Incident data were recorded by California Highway Patrol.

Download

The dataset is available via Tresorit: download [170 MB]

(Note: You have to provide your email address to download the dataset)

Content of the dataset

incidents.csv: Contains the collected and preprocessed 452 traffic incidents.

  • incident_id: Identifier of the incident (inherited from PeMS)
  • cc_code: CHP code (inherited from PeMS)
  • timestamp: The start of the incident (inherited from PeMS)
  • description: The description of the incident (inherited from PeMS)
  • latitude: Latitude coordinate of incident (inherited from PeMS)
  • longitude: Longitude coordinate of incident (inherited from PeMS)
  • district: Disctrict where the incident happened (inherited from PeMS)
  • freeway_nmbr: Freeway number (inherited from PeMS)
  • freeway_direction: Freeway direction (inherited from PeMS)
  • state_postmile: Relative postmile in California state: https://postmile.dot.ca.gov/PMQT/PostmileQueryTool.html?# (inherited from PeMS)
  • abs_postmile: Absolute postmile value (inherited from PeMS)
  • duration: Duration of the incident in minutes (inherited from PeMS)
  • up_id: Identifier of the upstream detector
  • up_lanes: Number of lanes observed by the upstream detector
  • up_distance: Distance between the incident and the upstream detector in miles
  • up_latitude: Latitude coordinate of upstream detector
  • up_longitude: Longitude coordinate of upstream detector
  • down_id: Identifier of the downstream detector
  • down_lanes: Number of lanes observed by the downstream detector
  • down_distance: Distance between the incident and the downstream detector in miles
  • down_latitude: Latitude coordinate of downstream detector
  • down_longitude: Longitude coordinate of downstream detector
  • active_onramp: Is there an active onramp between upstream and downstream detectors
  • active_offramp: Is there an active offramp between upstream and downstream detectors

To understand the postmile system in California please read the following article: https://www.cahighways.org/num-postmiles.html (It contains the explanation of negative detector distances)

traffic_data.parquet: Contains 5-minute aggregation traffic data for all traffic detector van be found in incidents.csv. (To work with parquet files, you need to install pyarrow package if you work with pandas)

The file contains the following columns:

  • timestamp: The timestamp of the measurement
  • speed: Measured average speed
  • occ: Measured average occupancy
  • flow: Measured total flow
  • station_id: Identifier of the detector

Citing

Please cite this article, if you use the dataset in your work. Bibitem and links to paper will be available soon.

License

The dataset is licensed under CC BY-NC 2.0.

license logo