recovery_study

Recovery Study

This directory contains scripts related to our prerpint:

@misc{weigend2021hydraulic,
      title={A hydraulic model outperforms work-balance models for predicting recovery kinetics from intermittent exercise}, 
      author={Fabian C. Weigend and David C. Clarke and Oliver Obst and Jason Siegler},
      year={2021},
      eprint={2108.04510},
      archivePrefix={arXiv},
      primaryClass={cs.OH}
}

Three types of scripts are available:

Compare models to data or to each other

Scripts that have a name that starts with compare_[last name]_dataset recreate comparison plots of the paper. You may use them to investigate the data we extracted from other studies or to see examples for how to use pypermod agents to predict recovery ratios.

compare_all recreates the big comparison table from the paper.

comparison_significance_test tests for statistical differences in prediction error distributions of two models as done in our paper.

compare_p_work_effec recreates the plot 4.1 of our paper paper

Fitting Tau

Scripts that have a name that starts with fitting recreate the fitting process of time constants for W'bal-weig and Chidnok comparisons in our paper. You may use them to further investigate our approaches to derive these time constants and how fitted models perform.

Simulate

note: the simulate script has been moved to the directory above as it is relevant for all data.

Scripts that have a name that starts with simulate use one or several models to simulate energy dynamics of an athlete during exercise. Use these scripts as examples for how to use pypermod for predictions.