This update is a significant revamp to our approach to the computing and reporting of growth. Instead of a single index, we fit a Bayesian GLM with MCMC that estimates exponential growth coefficients for each lineage based on geographically stratified frequency data. The output table is sorted by the 2.5% quantile of these values, ensuring that lineages that appear most likely to be exponentially growing and are of interest appear at the top.
This update is a significant revamp to our approach to the computing and reporting of growth. Instead of a single index, we fit a Bayesian GLM with MCMC that estimates exponential growth coefficients for each lineage based on geographically stratified frequency data. The output table is sorted by the 2.5% quantile of these values, ensuring that lineages that appear most likely to be exponentially growing and are of interest appear at the top.