Please cite our paper if you find our work useful for your research:

@article{zhang2022beyond,
  title={Beyond the limits of predictability in human mobility prediction: context-transition predictability},
  author={Zhang, Chao and Zhao, Kai and Chen, Meng},
  journal={IEEE Transactions on Knowledge and Data Engineering},
  year={2022},
  publisher={IEEE}
}

contextTransitionPredictability.py is the file that calculates context transition predictability based on the input file.
entropyCompute.py aims to calculate the relevant entropy and predictability.

The input data format is: [userid, locatonid_1@timestamp_1, ... , locatonid_n@timestamp_n].
The output file format is: [userid, entropy or predictability].

Input content(a partial trajectory extracted from the dataset):
1462,4b19f917f964a520abe623e3#Train Station@1333529414000,
4b19f917f964a520abe623e3#Train Station@1333549777000,
4b283516f964a520e19024e3#Train Station@1333550869000,
4b19f917f964a520abe623e3#Train Station@1333565041000,
4bbac8b753649c742f7249fb#Office@1333568100000,
4b600990f964a520e3d329e3#Train Station@1333568978000,
4b4aa852f964a520678c26e3#Train Station@1333614770000,
4b283516f964a520e19024e3#Train Station@1333615697000,
4b283516f964a520e19024e3#Train Station@1333647322000,
4b55c776f964a520a2ef27e3#Train Station@1333873221000,
4b4a79ccf964a520bc8826e3#Train Station@1333873488000

Context Transition Predictability output given temporal information:
1462,0.885632226078601