Overview

Repository is part of Vaccines Planning Tool for the COVID-19 Modeling Accelerator

For work supported by: Johns Hopkins University, the Society for Medical Decision Making, and the Rockefeller Foundation

This software is vaccine and variant of concern incorporated version of the following repository: https://github.com/haoxiangyang89/COVID_Staged_Alert

To learn more about the project, previous and on-going work:

• https://covid-19.tacc.utexas.edu/austin_covid_alert_system_-_voc.pdf
• https://www.nature.com/articles/s41467-021-23989-x
• https://www.pnas.org/content/117/33/19873.short

Installation and how to run the code

• Download and unzip the code to a local path (e.g., .../COVID19-vaccine-main/VaccineAllocation)

• The following packages are required:

  • matplotlib
  • pandas
  • numpy
  • Scipy

• Add both /COVID19-vaccine and /COVID19-vaccine/VaccineAllocation to your $PYTHONPATH

Running the Code

• Type the following code into Python Command Window before running any code -Import sys -sys.path.append(.../COVID19-vaccine-main) -sys.path.append(.../COVID19-vaccine-main/VaccineAllocation)

Structure of the Code

Config Directory

• Contains configuration files necessary to initialize running a sample path

Data Processing directory

• Contains death data, hospitalization, code cleaning, and seed generation files

Instances directory

• Contains related input .csv and .json files such as Omicron prevalence and hospitalization data

Output directory

• Used to store output files from Crunch. These files will be used in seed generation or plot generation

main_allocation.py

• The main module to run the simulation.
• Running a sample path in main_allocation.py
  • runfile('... /VaccineAllocation/main_allocation.py', 'austin -f setup_data_Final.json -t tiers5_opt_Final.json -train_reps 0 -test_reps 1 -f_config austin_test_IHT.json -n_proc 1 -tr transmission_new.csv -hos austin_real_hosp_updated.csv -v_allocation vaccine_allocation_fixed.csv -seed new_seed_Nov.p -n_policy 7 -v_boost=booster_allocation_fixed_50.csv -gt [-1,5,15,30,50]', wdir='.../VaccineAllocation')

Pipelinemultitier.py

• Responsible for generating plots after creating output files.

policy_search_functions.py

• Run the policy search on training set, if a trigger policy is not given. Perform test simulation on the best found policy.
• If a trigger policy is given perform test simulation for the given policy.

Epi_params.py

• Contains epidemiological parameters.

Interventions.py

• Defines the knobs of an interventions and forms the available interventions considering school closures, cocooning, and different levels of social distancing

SEIYAHRD.py

• Simulate the SEIR model with vaccines included, considering different age groups and seven compartments

Trigger_policies.py

• Different trigger policies that are simulated

Utils.py

• Timing and rounding functions

Vaccine_params.py

• Defines epidemiological characteristics and includes supply and fixed allocation schedule of vaccine

Vaccine_policies.py

• Includes different vaccine allocation policies that are simulated

Least_squares_fit.py

• Minimizes a weighted sum of least-square errors to fit transmission-reduction parameters and certain dynamics in use of the ICU and hospital duration

Init.py

ACS_script.py

main_ACS.py

Objective_functions.py

Threshold_policy.py

Output files

• The .p file (data file) will be generated in /output
• Generate plots using pipelinemultitier.py or generate seeds using seeds_read.py in the data processing directory