Code for simulating the metapopulation SEIR model. Sample network for US national simulation included.
A preliminary description of this model appeared in IEEE ICHI 2017: https://ieeexplore.ieee.org/document/8031141/
Journal version: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1007111 Cite as: Venkatramanan S, Chen J, Fadikar A, Gupta S, Higdon D, Lewis B, Marathe M, Mortveit H, Vullikanti A. Optimizing spatial allocation of seasonal influenza vaccine under temporal constraints. PLoS computational biology. 2019 Sep 16;15(9):e1007111.
Also used in upcoming paper: Venkatramanan S, Sadilek A, Fadikar A, et al. Forecasting influenza activity using machine-learned mobility map, to appear in Nature Communications.
The description of the model and a software manual can be found inside the doc/ folder.
PatchSim is compatible with Python 3.5+. It requires numpy and pandas.
Please use the different test*.py scripts to test the different functionalities of PatchSim. For more details on these features please check the software manual.
To test the US national simulation, extract the zip file test/US_county.zip and run "python test_det_US.py" from inside test/ folder.
Install CuPy version compatible with your CUDA version. For example: pip install cupy-cuda11x
Before importing PatchSim set PATCHSIM_USE_GPU
environment variable. For example:
- shell:
export PATCHSIM_USE_GPU=1
, before running PatchSim script - jupyter notebook:
%env PATCHSIM_USE_GPU=1
, before PatchSim import - python script:
import os; os.environ['PATCHSIM_USE_GPU'] = '1'
, before PatchSim import