Get epidemic phases from R(t)

This function loads in the posterior estimate of the retrospective R(t) from NNH. Then it categorizes each week into an epidemic phase based on the algorithm used in https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1011200 #nolint in the S5 appendix. Code available here: https://github.com/cdcepi/Evaluation-of-case-forecasts-submitted-to-COVID19-Forecast-Hub/blob/b3c379dfd48e8c673f67996014f151ce44cbd8fa/Code/Supplement%205_Rt_Epi%20Phases.R #nolint

Usage

get_epidemic_phases_from_rt(
  locations,
  retro_rt_path,
  location_col_name = "state_abb",
  prob_threshold = 0.9
)

Arguments

locations

A vector of the state abbreviations

retro_rt_path

A path to the parquet file of retrospective R(t) estimates

location_col_name

A string indicating the name of the column indicating location, default is state_abb

prob_threshold

A numeric between 0 and 1 that defines the threshold probability R(t) >= 1 or R(t) <= 1. Default is 0.9 from the above paper.

Value

df_epi_phase A dataframe containing the epidemic phase expanded to be daily, for each location, based on the retrospective R(t) estimate.