R/read.R

In BradyAJohnston/reluxr: Deconvolute Luminescence Readings on Bacterial Culture Plates

#' Logical test for well ID format.
#'
#' @param x A string vector.
#'
#' @return A logical vector.
#' @export
#'
#' @examples
#' is_well_id(c("a12", "a2", "a02", "foo1"))
is_well_id <- function(x) {
  well_id_regex <- "[:alpha:]{1,2}\\d{1,3}"

  stringr::str_detect(
    stringr::str_trim(x),
    well_id_regex
  )
}

#' Read the output of Tecan Plate Readers
#'
#' @param file File path to the `.xlsx` or `.csv` file.
#' @param temp Logical, whether to include the temperature column.
#'
#' @return a `tibble::tibble()` of the values from the file.
#' @export
#' @importFrom rlang .data
#'
#' @examples
#' fl <- system.file(
#'   "extdata",
#'   "calibrate_tecan",
#'   "calTecan1.xlsx",
#'   package = "reluxr"
#' )
#'
#' dat <- plate_read_tecan(fl)
#'
#' dat
plate_read_tecan <- function(file, temp = FALSE) {
  dat <- suppressMessages({
    readxl::read_excel(file, col_names = FALSE, .name_repair = "unique")
  })
  dat <- janitor::clean_names(dat)

  vec <- stringr::str_detect(dplyr::pull(dat, .data$x1), "Cycle Nr")
  vec <- dplyr::if_else(is.na(vec), FALSE, vec)
  vec <- cumsum(vec)
  vec <- c(vec[-1], max(vec))

  dat <- dat |>
    dplyr::mutate(
      chunk = vec,
    ) |>
    tidyr::nest(.by = "chunk") |>
    dplyr::filter(.data$chunk != 0) |>
    dplyr::mutate(
      signal = purrr::map_chr(.data$data, \(x) x[1, 1, drop = TRUE]),
      data = purrr::map(.data$data, dplyr::slice, -1),
      data = purrr::map(.data$data, janitor::row_to_names, row_number = 1)
    ) |>
    tidyr::unnest("data") |>
    janitor::clean_names() |>
    tidyr::drop_na("cycle_nr")

  dat <- dat |>
    tidyr::pivot_longer(
      cols = which(is_well_id(colnames(dat))),
      names_to = "well",
      names_transform = well_format,
      values_transform = as.numeric,
    ) |>
    tidyr::drop_na("value") |>
    dplyr::select(-"chunk") |>
    dplyr::mutate(
      dplyr::across(
        dplyr::matches("cycle_nr|time_s|temp"),
        as.numeric
      ),
      signal = stringr::str_extract(.data$signal, ".+(?=:)")
    ) |>
    dplyr::ungroup()

  if (temp) {
    dat
  } else {
    dat <- dplyr::select(dat, -dplyr::matches("temp"))
    dat
  }
}