Open yingycl opened 3 months ago
Thanks for your attention. We have updated the related code.
Thanks for the quick reply. Now I encounter another problem. In multi_main.py line 226: road_condition = np.load(condition_file) # T, N, N for i in range(road_condition.shape[0]): maxn = road_condition[i].max() road_condition[i] = road_condition[i] / maxn
I got maxn = 0 in most cases. Therefore, road_condition contained many nan values and finally the program terminated unexpectedly at
Traceback (most recent call last):
File "/MM-STGED/multi_main.py", line 340, in
We only provide example data, so the calculated road conditions are incomplete, and many values are zero. We have provided the final flow.npy in ./data/Porto/extra_data. You can use it directly and skip the step for calculating road conditions.
Problem solved. Thanks! I finished training on the example data, but the test accuracy is only 0.16. Is this because there is not enough training data? If so, where should I get the complete data?
python build_graph.py
Traceback (most recent call last): File "/MM-STGED/build_graph.py", line 259, in
G = build_global_POI_checkin_graph(dst_dir, new2raw_rid, raw_rn_dict, mbr, grid_size)
File "/MM-STGED/build_graph.py", line 68, in build_global_POI_checkin_graph
trajs = parser.parse(os.path.join(dst_dir, filename))
TypeError: parse() missing 1 required positional argument: 'id_dict'