/COVID19CaseNumberModel

Fit the algebraic growth of case numbers in Mainland China with an SIR-containment model

Primary LanguagePython

DOI

COVID-19 case number growth

The growth of case numbers concerning the recent COVID-19 outbreak in provinces of Mainland China can be modeled by a new SIR containment model. This is a complimentary repository that contains the data and the analysis discussed in the paper B. F. Maier and D. Brockmann, "Effective containment explains sub-exponential growth in confirmed cases of recent COVID-19 outbreak in Mainland China", 2020.

Data

The json-file data/all_confirmed_cases_with_population.json contains case number data of the currently affected provinces in China as well as population size.

The time series count the aggregate number of people whose infection was laboratory-confirmed. It was gathered by the Johns Hopkins University Center for Systems Science and Engineering.

For the data contained in mainland_china, all province data except the one from Hubei was aggregated by means of interpolation.

Since Feb 12 the case data includes symptomatic cases without lab-confirmation, as well, therefore we only consider data from before Feb 12 6am.

Prerequisites

Written and tested for Python 3.7

Requirements

pip install requirements.txt

These are the requirements:

simplejson==3.16.0
numpy==1.17.2
scipy==1.3.1
bfmplot==0.0.7
lmfit==0.9.12
tabulate==0.8.2
matplotlib==3.0.2
tqdm==4.28.1

Examples

Reproduce plots

cd main_results
python model_large_hubei_and_mainland_china.py fit_parameters/hubei_china.p
python model_fit_confirmed_cases_500.py fit_parameters/confirmed_cases_500.p

modelFitHubeiMainland

modelFitConfirmed500

In case you want new fits, do

python model_large_hubei_and_mainland_china.py
python model_fit_confirmed_cases_500.py

The fit parameters are saved in main/results/fit_parameters/confirmed_cases_500.p

Works similarly for the other analysis scripts.