R/cumulative-sequence.R

In poissonconsulting/poisdata: Read and Write Data

#' Cumulative Sequence
#'
#' Determines when in sequence a cumulative sum is reached.
#' Useful for ATUs.
#'
#' @param x A scalar specifying the start.
#' @param cumsum A scalar specifying the cumulative sum.
#' @param data The data.frame.
#' @param sequence A string naming the sequence column.
#' @param value A string of the value column.
#' @return A scalar of the end.
#' @export
ps_cumulative_sequence <- function(x, cumsum, data, sequence = "DateTime", value = "Value") {
  check_dim(x, values = 1)
  check_dim(cumsum, values = 1)
  chk_string(sequence)
  chk_string(value)

  check_data(data)
  check_names(data, sequence)
  check_names(data, value)


  if (sequence == value) {
    ps_error("value column '", value, "' must not be the same as sequence")
  }

  data %<>%
    tibble::as_tibble() %>%
    dplyr::arrange(UQ(parse_quo(sequence, env = rlang::caller_env())))

  na <- x
  is.na(na) <- TRUE

  if (nrow(data) < 2L) {
    return(na)
  }

  if (!identical(length(unique(diff(data[[sequence]]))), 1L)) {
    ps_error("sequence must be unique and complete (try ps_add_missing_sequence)")
  }

  if (!x %in% data[[sequence]]) {
    return(na)
  }

  data <- data[data[[sequence]] > x, ]
  data$cumsum <- cumsum(data[[value]])
  data <- data[!is.na(data$cumsum) & data$cumsum >= cumsum, ]

  if (!nrow(data)) {
    return(na)
  }

  data[[sequence]][1]
}