PSLmodels / COVID-MCS

COVID-MCS

Introduction

COVID-MCS is an web application that allows users to apply the apply the testing framework in Ganz (2020) to data on positive test rates. See Ganz (2020) for more details on the implementation of the test.

Why COVID-MCS?

Policymakers and public health researchers have had difficulty determining whether benchmarks surrounding trends in the intensity of the COVID-19 pandemic have been satisfied. One reason for the confusion is that commonly-used hypothesis testing methods, e.g., t-tests for mean comparisons or linear regression models, are poorly suited for questions about shapes of trends in data, e.g., sustained decrease.

The hypothesis testing framework implemented here is suitable to ask questions like “has a region experienced a specified number of days of declining COVID-19 cases?” by applying the model confidence set (MCS) framework to shape-constrained regression analysis. Based on the output of the test, the analyst can determine whether the data are consistent with a region's gating criteria.

Applying the test to your own data

Analyze data from your web browser

• The easiest way to apply this methodology is to use the web application hosted on Compute Studio.

Analyze data using your own computer

• The methodology can be used on your personal computer by using either R or Python.

Installing COVID-MCS

• To install COVID-MCS, run the following in the folder you wish to store the files:

git clone https://github.com/PSLmodels/COVID-MCS
cd COVID-MCS
pip install -e .

Using R

• All necessary functions can be called from COVID-MCS/main.R. View a summary of the test by called summary(m).

Using Python

• Modify the default values defined in COVID-MCS/defaults.json in COVID-MCS/adjustment_file.json.
• Create an instance of the COVID_MCS_TEST class by calling c = COVID_MCS_TEST().
• Return the summary text and graphs by calling summary, graphs = c.MCS_Test()