Skip to content

Latest commit

 

History

History
88 lines (71 loc) · 4.48 KB

README.md

File metadata and controls

88 lines (71 loc) · 4.48 KB

My stupid attempt at modeling the coronavirus outbreak

Author : Ali Snedden

License: MIT

Purpose:

This is a really ignorant attempt at understanding covid19. I have no experience in these type of simulations or analyzing this kind of data. This repos contains programs that do modeling and plotting

Installation and Dependencies :

  1. Python 3 installation that includes common libraries such as numpy and matplotlib.
  2. ffmpeg - Only useful for generating movies from the simulations

Usage:

This repo does several things.

  1. It plots and fits the Itallian covid-19 data from the national government. E.g.
python3 ./src/plot_italy_data.py option_to_plot italy_data_file [log-lin] [index]
      option_to_plot :
           'hosp_w_sympt'  : Hospitalized w/ Symptoms
           'icu'           : Intensive Care Units
           'total_hosp'    : All Hospitalized
           'home_isolation': Home Isolation
           'total_positive': All Positive
           'new_positive'  : New Positive
           'discharged'    : Discharged
           'dead'          : Dead
           'total_cases'   : Total Cases
           'swabs'         : Swabs
      italy_data_file : the path to the dpc-covid19-ita-andamento-nazionale.csv
      [log-lin or lin-lin] : optional, plot y axis in natural log, if straight line
                             then experiencing exponential growth. If not specified,
                             log-lin assumed
      [slice_index]        : for fitting, e.g.
                             if = -10, it will fit the last 10 points
                             if = 10, it will fit the first 10 points
  1. It plots the country cases (see John Hopkins' data) as a function of time. E.g.
python3 ./src/plot_jhu_data.py country log-lin slice_index
      country   : See time_series_covid19_confirmed_global.csv
                  for coutries to plot options
      log-lin   : required, plot y axis in natural log, if fit is
                  straight line then experiencing exponential growth.
                  My hope is to someday implement other to be fit types
                  (e.g. lin-lin)
      slice_index : required, for fitting, e.g.
                  if = -10, it will fit the last 10 points
                  if = 10, it will fit the first 10 points
  1. It plots several countries' death data (see John Hopkins' data) showing 1 day and 3 day doubling times. Motivated by NYT plot. E.g.
python src/plot_country_deaths.py
  1. An Ising inspired model of infection. It is a 1D model. This was my first attempt at modeling infectious spread. With an infectious agent, it can infect any other agent in the 1D model. This non-local behavior is obviously realistic, yet is fun anyways.
python3 ./src/simulation_1D.py R0 inc_time infect_time
      R0          : How many people on average person infects
      inc_time    : Incubation time (days)
      infect_time : Time after infection person becomes infectious
  1. A fun little toy program that emulates the type of plotting / simulations done in Simulating an Epidemic - YouTube.
python3 ./src/simulating_agent_model_2D.py [quarantine]
   [quarantine] : optional, puts infected in quarantine symptoms present


   After running, create a movie via :
       ffmpeg -framerate 4 -pattern_type glob -i 'tmp/*.png' -c:v libx264 out.mp4

References:

  1. Strategies for mitigating an influenza pandemic
  2. Impact of non-pharmaceutical interventions (NPIs) to reduce COVID- 19 mortality and healthcare demand
  3. Italian Department of Civil Protection
  4. Institute for Health Metrics and Evaluation
  5. Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship
  6. Simulating an Epidemic - YouTube