libcity.data.dataset.trajectory_encoder.serm_encoder

class libcity.data.dataset.trajectory_encoder.serm_encoder.SermEncoder(config)[source]

Bases: libcity.data.dataset.trajectory_encoder.abstract_trajectory_encoder.AbstractTrajectoryEncoder

encode(uid, trajectories, negative_sample=None)[source]

standard encoder use the same method as DeepMove

Recode poi id. Encode timestamp with its hour.

Parameters

  • uid ([type]) – same as AbstractTrajectoryEncoder

  • trajectories ([type]) –

    same as AbstractTrajectoryEncoder trajectory1 = [

    (location ID, timestamp, timezone_offset_in_minutes), (location ID, timestamp, timezone_offset_in_minutes), …..

    ]

gen_data_feature()[source]
get_text_from_point(point)[source]

return word index

load_wordvec(vecpath='./raw_data/word_vec/glove.twitter.27B.50d.txt')[source]