Process Cholera Data to Calculate Aggregated Case Fatality Ratios and Fit Beta Distributions
Source:R/process_CFR_data.R
process_CFR_data.RdThis function processes WHO annual cholera data, aggregates cases and deaths for each country from 2014 through the last complete year in the data, calculates the aggregated Case Fatality Ratio (CFR), and fits Beta distributions for each country. Countries with fewer cases than min_obs will have their CFR set to the value of the AFRO Region.
Arguments
- PATHS
A list containing the paths where the cholera data is located. Typically generated by the
get_paths()function and should include:DATA_WHO_ANNUAL: Path to the directory containing processed WHO annual cholera data.
DOCS_TABLES: Path to the directory where processed CFR data will be saved.
- min_obs
An integer specifying the minimum number of cases required for a country to have its own CFR. Countries with fewer cases than this threshold will have their CFR set to the value of the AFRO Region.
Value
Invisibly, a data frame with aggregated CFR, confidence intervals, and Beta distribution parameters for each country (also written to case_fatality_ratio_2014_<end year>.csv in DOCS_TABLES and DATA_WHO_ANNUAL).
Details
A calendar year still in progress when its dashboard snapshot was taken is dropped before aggregating (its deaths lag its cases), using the same rule as est_CFR_hierarchical; the end year of the output (and of its filename) is the last year left after that filter. A country whose only data fall in an in-progress year therefore gets the AFRO Region CFR. Any existing case_fatality_ratio_2014_<year>.csv with a later end year (one built while that year was still counted) is deleted from both output directories, so plot_CFR_by_country, which reads the highest end year, picks up the new file. Country-years with a missing case or death count are left out of both totals, so each CFR is computed over the same years for its numerator and denominator. Rows are aggregated by ISO code, and each country is labelled with its most frequent spelling in the data (the WHO file spells some countries more than one way).
Examples
if (FALSE) { # \dontrun{
PATHS <- get_paths()
cholera_data <- process_CFR_data(PATHS, min_obs = 100)
} # }