Compartmental Epidomiology Modeling

py0 is a python implementation of compartmental disease modeling.

Installation

To install py0:

pip install py0@git+https://github.com/ur-whitelab/py0.git

Maximum Entropy Biasing

py0 can be coupled with MaxEnt to modify epidomiology parameters to find the best fit to disease trajectory given a set of observations and also infer the true origin of the outbreak (patient-zero). These observations are time-averaged fractional values that can come from different compartments (S, E, A, I and R) of a known synthetic reference trajectory or real pandemic spread data.

Creating an Ensemble of Trajectories

We try to explore the disease trajectory space over a distribution of epidomiology parameters, while changing the infection origin to different nodes (counties).

MaxEnt Fit

MaxEnt Installation

The package uses Keras (Tensorflow). To install:

pip install maxent-infer

Citation

See paper and the citation:

@article{ansari2022inferring,
  title={Inferring spatial source of disease outbreaks using maximum entropy},
  author={Ansari, Mehrad and Soriano-Pa{\~n}os, David and Ghoshal, Gourab and White, Andrew D},
  journal={Physical Review E},
  volume={106},
  number={1},
  pages={014306},
  year={2022},
  publisher={APS}
}

License

Authors

py0 is developed by Mehrad Ansari, Rainier Barrett and Andrew White.