This is an urban traffic speed dataset which consists of 214 anonymous road segments within two months (i.e., from August 1, 2016 to September 30, 2016) at 10-minute interval, and the speed observations were collected from Guangzhou, China. A detailed description and files of this dataset are also available at Zenodo -- Urban Traffic Speed Dataset of Guangzhou, China.
speeddata.csv - the traffic speed dataset (contains 1,855,589 speed observations). Note that, for the convenience, speeddata.csv is separated into two files (i.e., speeddata_Aug.csv & speeddata_Sep.csv) where speeddata_Aug.csv covers the total observations during August, 2016 and speeddata_Sep.csv covers the total observations during September, 2016, respectively.
(1) road_id: a unique anonymous identifier for each road segment. As an example, 1 indicates the first road segment;
(2) day_id: a unique code indicating the date. In this column, 1 represents Aug. 1, 2016, 2 represents Aug. 2, 2016, as such, 61 represents Sep. 30, 2016;
(3) time_id: a unique code indicating the time windows. For example, 1 represents 00:00:00-00:10:00, 2 represents 00:10:00-00:20:00;
(4) speed: the speed values with unit km/h.
tensor.mat - the third-order tensor in Matlab and it can be directly loaded. In detail, we have
(1) the length of its first dimension corresponding to road semgent is 214;
(2) the length of second dimension corresponding to day is 61;
(3) the length of third dimension corresponding to time window is 144;
(4) the tensor entries is valued by traffic speed where 0 indicates the unobserved.
The number of non-zero entries of this tensor is 1,855,589 and the total entries is 1,879,776. So, the missing rate of this tensor is originally given by 1.29%.
Xinyu Chen, Zhaocheng He, Jiawei Wang, 2018. Spatial-temporal traffic speed patterns discovery and incomplete data recovery via SVD-combined tensor decomposition. Transportation Research Part C: Emerging Technologies, 86, 59-77. Download PDF