Code to estimate SARS-CoV-2 incidence based on changing numbers of diagnostic tests. Accompanies the manuscript: "Disentangling Increased Testing from Covid-19 Epidemic Spread". A blog post describing details of this project can be found here.
This model uses a single free parameter c to model the relative propensity of testing SARS-CoV-2-infected individuals compared to non-infected individuals. We use two strategies to estimate this free parameter c:
- Seroprevalence studies
- Symptomatic Rates
There are two sets of experiments: estimates of daily incidence can be found in 'estimate_daily_incidence.ipynb', while estimates of cumulative incidence can be found in estimate_total_cases.ipynb. These notebooks automatically download new data every day, so you can re-run if the results in this repository are out of date.
A few representative plots are shown below; results for all 50 states can be found in the results/ directory.
