covid-schools

Analyses for school reopening. The folder first_analysis includes the folders and analyses that were used for the initial paper submission:

https://www.medrxiv.org/content/10.1101/2020.09.08.20190942v1

The remaining folders are in active use and development, especially testing_in_schools.

Requirements

Python >=3.6 (64-bit). (Note: Python 2 is not supported, and only Python >=3.8 has been tested.)

We also recommend, but do not require, using Python virtual environments.

Installation

Create a virtual environment, if desired:

conda create -n covaschool python=3.8
conda activate covaschool

Install Covasim:

pip install covasim --upgrade

Install SynthPops in a folder of your choice:

git clone https://github.com/InstituteforDiseaseModeling/synthpops
cd synthpops
python setup.py develop

Install CovasimSchools:

git clone https://github.com/amath-idm/covid-schools
cd covid-schools
python setup.py develop

Quick start

Run the main test script:

cd tests
python test_schools.py

If it worked, it should bring up a plot.

Slower start

To run the main analysis scripts, first generate the synthetic populations:

cd testing_in_schools
python create_sp_pop.py

Then run either run_scenarios.py or sensitivity_scenarios.py. NB, these are intended to be run on HPCs and will likely not work (or take more than a day to complete) on a PC.

Fonts

The font "Roboto Condensed" is used for generating the figures. To use with Python:

On Linux/Mac:

1. Copy the font files you want (e.g. in the assets folder) to $HOME/.fonts or /usr/share/fonts/truetype/
2. Run sudo fc-cache -f -v
3. Confirm they're installed: fc-list

On Windows:

1. Open the assets folder in Windows Explorer and do what you need to do.

Then on all systems:

1. Rebuild matplotlib font cache: from matplotlib import font_manager as fm; fm._rebuild() -- this needs to be done for every user and/or virtual environment
2. Confirm they're available: from matplotlib import font_manager as fm; fm.findSystemFonts()