automutualinformation 0.1.2

Auto Mutual Information (Sequential Mutual Information) for temporal data.

autoMI

Auto Mutual Information (Sequential Mutual Information) for temporal data.

Auto mutual information can be treated as the equivalent of autocorrelation for symbolic data.

For more info references see:

• Mutual information functions versus correlation functions. W Li. (1990). Journal of Statistical Physics
• Critical Behavior in Physics and Probabilistic Formal Languages. HW Lin, M Tegmark (2017) Entropy
• Parallels in the sequential organization of birdsong and human speech. T Sainburg, B Thielman, M Thielk, TQ Gentner, (2019) Nature Communications
• Long-range sequential dependencies precede complex syntactic production in language acquisition. T Sainburg, A Mai, TQ Gentner. Proceedings of the Royal Society B

Installation

The python package is installable via pip.

pip install automi

Quick Start

from auto_mi import sequential_mutual_information as smi
(MI, MI_var), (shuff_MI, shuff_MI_var) = smi(
    [signal], distances=range(1,100)
)

Documentation

Documentation and usage information is currently available in jupyter notebooks in the notebooks folder.

Citation

If you use this package, please cite the following paper:

@article {NBC2020,
    author = {Sainburg, Tim and Mai, Anna and Gentner, Timothy Q.},
    title = {Long-range sequential dependencies precede complex syntactic production in language acquisition},
    journal = {Proceedings of the Royal Society B},
    doi = {https://dx.doi.org/10.1098/rspb.2021.2657},
    year = 2022,
    }

TODO

• make pypi package
• create tests/travisci
• add additional parameters example