Context This dataset is a combination between the [NYC Taxi duration challenge][1] and the [DIMACS][2] shortest paths dataset. Content This set contains three files: Each intersection in NY gets an id as well as GPS coordinates in USA-road-d.NY.co It looks like v 1 -73530767 41085396 v 2 -73530538 41086098 v 3 -73519366 41048796 v 4 -73519377 41048654 v stands for vertex and 1-264346 and the longitude -73530767 represents 73.530767 W whereas the next column here: 41085396 holds the information about the latitude in the same format. The second file "USA-road-d.NY.gr" contains the information about how the intersections mentioned in the previous file are connected and the travel distance between them. (Unfortunately I don't know the unit) a 1 2 803 a 2 1 803 a 3 4 158 a 4 3 158 a 5 6 774 a is always the first character for the data in this file and stands for "arrow". The next column represents the id of the start vertex and the second one the end vertex. The last column gives us the information about the distance between those two vertices. The last part of this set is the combination of the data mentioned above and the [NYC Taxi duration challenge][1] dataset. I just add two columns s_id and t_id where s is the start and t is the target. The dropoff and the pickup location are mapped to the id of the nearest vertex in this dataset. Acknowledgements This data about the vertices is directly given from [DIMACS][2]. Thanks for the data which is extremely helpful for visualizing the data given in the challenge. Inspiration This data was generated for the challenge about taxis in NYC and used for the visualization. You can use it for every project which needs some insight about the street structure of NYC. [1]: https://www.kaggle.com/c/nyc-taxi-trip-duration [2]: http://www.diag.uniroma1.it/challenge9/download.shtml