sscollis / sanseir

Solve a stochastic SEIR model for disease progression

SanSEIR

Solves a stochastic SEIR model for disease progression. SanSEIR reproduces the results presented in the paper

Yao Yu Yeo, Yao-Rui Yeo, Wan-Jin Yeo, A Computational Model for Estimating the Progression of the COVID-19 Cases in the US West and East Coasts, MedRxiv, https://doi.org/10.1101/2020.03.24.20043026

Sample results based on the inputs of Yeo et al. are given in the yeo-etal directory.

Building and Running

To build with gfortran simply:

make

You will need to modify the Makefile if you use a different FORTRAN compiler. Once you have sanseir build, to run simply pass the input file (in FORTRAN namelist format) as an argument

cd yeo-etal/usa
../../sanseir.exe usa.inp

This generates an ensemble of 50 runs with output files called output.*, scaled.*, rates.*.

These can be plotted using gnuplot. For example, to plot the cummulative deaths

gnuplot
plot for [i=1:50] file=sprintf("output.%d",i) file using 1:12 w l lt 'grey' title ""

There are a number of gnuplot scripts for making typical plots in the plot directory.

For example, to make the West Coast plot shown at the bottom of this page, do the following

cd yeo-etal/west
../../sanseir.exe west.inp
gnuplot
load "../../plot/all-infected.com"

Descriptions of contents

Item	Description
yeo-etal	Inputs and results for https://doi.org/10.1101/2020.03.24.20043026
test	Additional experimental input files
plot	Scripts to make plots using gnuplot
sanseir.f90	SanSEIR source code in FORTRAN90
cleanup	Script to cleanup all output files from SanSEIR`

Baseline results: Infections on the West Coast

Baseline results: Infections on the East Coast

Baseline results: Infections for USA

S. Scott Collis\