Covid-19-Model-Validation

Description

Source code for our Medium article:

Transparency, Reproducibility, and Validity of COVID-19 Projection Models

Contents

IHME

Validation on Italy

1. Create training data   IHME/Italy/Data_creation/IHME_data_creation.R
2. IHME model fitting   IHME/Italy/Model_validation/model_validation_Italy.py
3. Calcuate confidence interval for the predicted cumulative death   IHME/Italy/Prediction_interval_calculation/IHME_nystate_prediction_confidence_interval_Italy.R  

Validation on New York State

1. Create training data   IHME/NYstate/Data_creation/IHME_data_creation.R
2. IHME model fitting   IHME/NYstate/Model_validation/model_validation_nystate.py
3. Calcuate confidence interval for the predicted cumulative death   IHME/NYstate/Prediction_interval_calculation/IHME_nystate_prediction_confidence_interval_NY.R 

Plots

Generate input for Tableau to create the IHME figures in the article   IHME/Plots/Create_input_data_for_plots.R  

ICL

Detailed documentation for ICL is provided at ICL/README.md.

eSIR

Validation on Italy

eSIR/Italy/Validation_Italy.R

Validation on New York State

eSIR/NYstate/Validation_NYstate.R

Plots

Generate input for Tableau to create the eSIR figures in the article 
   -- Italy: eSIR/Plots/Create_input_data_for_tableau_figures_Italy.R
   -- New York State: eSIR/Plots/Create_input_data_for_tableau_figures_NYstate.R

Additional Information

IHME (Institute of Health Metrics at the University of Washington)

• Homepage: https://covid19.healthdata.org/

• Code: https://github.com/ihmeuw-msca/CurveFit

• References:

  COVID, I. and Murray, C.J., 2020. Forecasting the impact of the first wave of the COVID-19 pandemic on hospital demand and deaths for the USA and European Economic Area countries. medRxiv.

• Accession date: April 17, 2020

ICL (Imperial College London)

• Homepage: https://www.imperial.ac.uk/about/covid-19/

• Code: https://github.com/ImperialCollegeLondon/covid19model

• References:

  Flaxman, S., Mishra, S., Gandy, A., Unwin, H., Coupland, H., Mellan, T., Zhu, H., Berah, T., Eaton, J., Perez Guzman, P. and Schmit, N., 2020. Report 13: Estimating the number of infections and the impact of non-pharmaceutical interventions on COVID-19 in 11 European countries.

  Flaxman, S., Mishra, S., Gandy, A., Unwin, H.J.T., Coupland, H., Mellan, T.A., Zhu, H., Berah, T., Eaton, J.W., Guzman, P.N. and Schmit, N., 2020. Estimating the number of infections and the impact of non-pharmaceutical interventions on COVID-19 in European countries: technical description update. arXiv preprint arXiv:2004.11342.

• Accession date: April 15, 2020

eSIR (University of Michigan)

• R package eSIR: https://github.com/lilywang1988/eSIR

• References:

  Song, P.X., Wang, L., Zhou, Y., He, J., Zhu, B., Wang, F., Tang, L. and Eisenberg, M., 2020. An epidemiological forecast model and software assessing interventions on COVID-19 epidemic in China. medRxiv.

  Ray, D., Salvatore, M., Bhattacharyya, R., Wang, L., Mohammed, S., Purkayastha, S., Halder, A., Rix, A., Barker, D., Kleinsasser, M. and Zhou, Y., 2020. Predictions, role of interventions and effects of a historic national lockdown in India's response to the COVID-19 pandemic: data science call to arms. medRxiv.

• Accession date: April 20, 2020

Data Sources

• Country-level data for Italy and India: JHU CSSE COVID-19 GitHub (Accession date: May 2nd, 2020)

• State-level data for New York: New York Times data (Accession date: May 2nd, 2020)

Contributors

• Neha Agarwala, Department of Mathematics and Statistics, University of Maryland, Baltimore County
• Jin Jin, Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health
• Prosenjit Kundu, Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health
• Yi Wang, Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health
• Ruzhang Zhao, Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health
• Nilanjan Chatterjee, Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health