R/read_kmi_data.R

In inbo/inborutils: Collection of Useful R Utilities

#' Load `KMI` data as data.frame
#'
#' @param filename path/filename of the `KMI` datafile
#' @param n_max integer; shortcut for testing, loading just a section of the
#' data
#'
#' @return `data.frame` with the headers `location_name`, `datetime` (UTC),
#' `value`, `unit`, `variable_name`, `latitude`, `longitude` and
#' `source_filename`
#'
#' @export
#' @importFrom readr read_delim cols col_character col_double col_integer
#' col_datetime
#' @importFrom dplyr select mutate %>%
#' @importFrom rlang .data
#' @importFrom lubridate tz
#'
#' @family download_functions
#'
#' @examples
#' \dontrun{
#' # see vignettes for more examples
#' file_path <- "kmi_file.txt"
#' precipation_kmi <- read_kmi_data(file_path)
#' }
read_kmi_data <- function(filename, n_max = Inf) {
  col_types <- cols(
    date = col_datetime("%Y-%m-%d_%H"),
    JAAR = col_integer(),
    MAAND = col_integer(),
    DAG = col_integer(),
    UUR = col_integer(),
    STATION = col_character(),
    `NEERSLAG(mm)` = col_double()
  )
  raw_data <- read_delim(filename,
    delim = ";",
    col_names = TRUE,
    col_types = col_types,
    n_max = n_max
  )

  # currently assumes that data is already UTC time zone (default of readr)
  # currently assumes KMI data is ALWAYS precipitation data...
  sprintf("Time zone of data is %s", tz(raw_data$date))
  kmi_data <- raw_data %>%
    select(
      datetime = .data$date,
      location_name = .data$STATION,
      value = .data$`NEERSLAG(mm)`
    ) %>%
    mutate(unit = "mm") %>%
    mutate(variable_name = "precipitation") %>%
    mutate(source_filename = basename(filename)) %>%
    mutate(quality_code = "")

  kmi_data
}