Traceback (most recent call last):
File "E:\codes\github\ST2Vec\preprocess.py", line 203, in
prepare_dataset(trajfile=str(config["traj_file"]), timefile=str(config["time_file"]), kseg=config["kseg"])
File "E:\codes\github\ST2Vec\preprocess.py", line 34, in prepare_dataset
node_list_int = np.array(node_list_int)
^^^^^^^^^^^^^^^^^^^^^^^
ValueError: setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (29830,) + inhomogeneous part.

After I use openStreetMap to get road network data(a shp file), what should I do to preprocess it so that I can get the experiment data?

Hi there. Sorry to bother again. First of all, I would like to express my gratitude for sharing your amazing work and responding the previous questions! I am trying to understand the codes for map matching part so I added the following codes to STmatching_distribution_ver.py. Hope it is correct by logic.

if __name__ == '__main__':
    import pickle
    import pandas
    fr = open("data/tdrive/st_traj/tdrive.pkl",'rb')
    trs = pickle.load(fr) 
    traj_df = pd.DataFrame([])
    for col, values in trs.iterrows():
        traj_i = pd.DataFrame(data=values['trajectory'], columns=['LON', 'LAT'])
        traj_i['TRAJ_ID'] = values['id']
        trajectory = values['trajectory']
        traj_df = pd.concat([traj_df, traj_i], axis=0)
    print("end")
    traj_task_list = data_convert(traj_df)
    mmtr_list = []
    for traj in traj_task_list:
        mmtr = trajectory_matching(traj)
        mmtr_list.append(mmtr)
    print(mmtr_list)

However, most matched vertex sequences in mmtr_list contain some nan vertices. Should I ignore nan values or just eliminate those trajectories?

Also, Could you explain a little bit more about how to get the ultimate matching_result.csv file? Maybe I'm missing something.

Thanks so much!

Zixin Qin
BJTU