R/unwrap_demographics.R

In ramiromagno/quincunx: REST API Client for the 'PGS' Catalog

unwrap_demographics <- function(tbl_json) {
  sample_age <- tbl_json |>
    tidyjson::enter_object('sample_age') |>
    unwrap_demographic() |>
    tibble::add_column(variable = 'age', .before = 'estimate_type')

  followup_time <- tbl_json |>
    tidyjson::enter_object('followup_time') |>
    unwrap_demographic() |>
    tibble::add_column(variable = 'follow_up_time', .before = 'estimate_type')

  demographics <-
    dplyr::bind_rows(sample_age, followup_time)

  return(demographics)

}

unwrap_demographic <- function(tbl_json) {
  tbl1 <-
    tbl_json |>
    tidyjson::spread_values(
      estimate_type = tidyjson::jstring('estimate_type'),
      # `estimate` is sometimes not a number but a string because it includes
      # the interval... sigh... Example: "62.4 [48.9, 75.9]".
      # estimate = tidyjson::jnumber('estimate'),
      unit = tidyjson::jstring('unit'),
      variability_type = tidyjson::jstring('variability_type'),
      variability = tidyjson::jnumber('variability')
    ) |>
    # Workaround to handle also the cases when `estimate` values come as
    # "62.4 [48.9, 75.9]".
    dplyr::mutate(estimate = parse_estimate(.data)) |>
    tidyjson::enter_object('interval') |>
    unwrap_interval()

  return(tbl1)
}