NYC-Dataset icon indicating copy to clipboard operation
NYC-Dataset copied to clipboard

Organize some grid-based traffic flow datasets, mainly New York City bicycle and taxi data

NYC Dataset

The traffic dataset of New York City is a commonly used dataset in traffic prediction problems, mainly including taxi traffic data and bicycle traffic data.

  • NYC Taxi: https://www1.nyc.gov/site/tlc/about/tlc-trip-record-data.page
  • NYC Bike: https://www.citibikenyc.com/system-data

Many papers in the field of traffic prediction (especially in the field of traffic flow prediction) use part of these two large-scale datasets.

Data List

Here I collected 8 relevant datasets of 4 papers, all of which are grid-based datasets. If you use these datasets, please cite the original paper.

[AAAI2017] Deep Spatio-Temporal Residual Networks for Citywide Crowd Flows Prediction

  • TaxiBJ (Range 20130601-20131030 / 20140301-20140630 / 20150301-20150630 / 20151101-20160410, Map size 32x32)
  • NYCBike (Range 20140401-20140930, Map size 16x8)

[ACM MM 2018] Attentive Crowd Flow Machines

  • NYCTaxi (Range 20140101-20141231, Map size 15x5)

[AAAI2019] Revisiting Spatial-Temporal Similarity: A Deep Learning Framework for Traffic Prediction

  • NYCTaxi (Range 20150101-20150301, Map size 10x20)
  • NYCBike (Range 20160701-20160829, Map size 10x20)

[KDD2019] Urban Traffic Prediction from Spatio-Temporal Data Using Deep Meta Learning

  • T-Drive (Range 20150201-20160602, Map size 32x32)

[KDD2020] Preserving Dynamic Attention for Long-Term Spatial-Temporal Prediction

  • NYCTaxi (Range 20160101-20160229, Map size 16x12)
  • NYCBike (Range 20160801-20160929, Map size 14x8)

Data Link

Please visit this link to download the dataset you need, the verification code is 1231.