R/checkUserED.R

In MolyWang/microCompet: Identify A Microbe's Potential Nutrition Competitor From Microbiota

#' Check Whether The User Input ED Dataframe Satisfies Function Requirement
#'
#' Given a data.frame ED and two user specified indices, check whether these arguments can
#' be used successfully by functions \code{overallSimilarity} and \code{competeMicrobiota}.
#' This function only check for common requirements, and each function carries out the
#' data.frame update and function-specific checks.
#' Function stops early if it identifies errors cannot be corrected by function.
#'
#' @param ED The data.frame to be passed in as ED arguments in other functions.
#' @param firstMicrobe The value to be passed in as firstMicrobe in other functions.
#' @param lastMicrobe The value to be passed in as lastMicrobe in other functions.
#'
#' @return A two-element logical vector. The first indicating whether lastMicrobe needs to
#'     be updated in respective function, while the second indicating whether the ED frame
#'     contains unexpected values (other than 0, 1, T/F may be accepted but not recommended)
#'     in columns between firstMicrobe and lastMicrobe.
#'
#' @examples
#'  \dontrun{
#'   reportVec <- checkUserED(EnzymeDistribution, 4, 10)
#'   reportVec
#'  }
#'
#' @export
#'
checkUserED <- function(ED, firstMicrobe, lastMicrobe) {
  # ============ Check for problems that have to "stop" ============
  # check user provided indices firstMicrobe and lastMicrobe are positive integers
  if (firstMicrobe <= 0 | floor(firstMicrobe) != firstMicrobe |
      lastMicrobe <= 0 | floor(lastMicrobe) != lastMicrobe) {
    stop("firstMicrobe or lastMicrobe is invalid, positive integers are required.")
  }

  # check all required columns are present with specified names
  if (!("Gene" %in% colnames(ED))) {
    errorInfo <- sprintf("Columns Gene (case sensitive) are required, and at least one is missing in your data.frame.")
    stop(errorInfo)
  }

  # check at least one genomes are provided in ED
  if (firstMicrobe > lastMicrobe) {
    stop("Invalid column indices, firstMicrobe should be no larger than lastMicobe.")
  }

  # check indices are no larger than total number of columns available
  if (firstMicrobe > ncol(ED)) {
    stop("Start index firstMicrobe should be no larger than total number of columns available. (", ncol(ED), ")")
  }

  # ============ Check for problems can be solved ============
  # initiate report vector
  reportVec <- c(FALSE, FALSE)
  names(reportVec) <- c("lastMicrobeTooLarge", "UnexpectedValuesInED")

  # Remind if lastMicrobe is larger than total columns available
  if (lastMicrobe > ncol(ED)) {
    sprintf("The given lastMicrobe index is larger than total columns available, and will be replaced by %d.",
            ncol(ED))
    reportVec["lastMicrobeTooLarge"] <- TRUE
  }

  # check whether columns from firstMicrbe to lastMicrobe contains only 0 and 1.
  if (lastMicrobe > ncol(ED)) {
    lastMicrobe <- ncol(ED)
  }
  relevantSection <- ED[firstMicrobe:lastMicrobe]
  unexpectedValues <- (relevantSection != 0) & (relevantSection != 1)
  unexpectedValuesCount <- sum(as.integer(unexpectedValues))
  if (unexpectedValuesCount > 0) {
    sprintf("%d cells from column %d to column %d in the provided data.frame is not 0 or 1, and will be replaced by 0.",
            sum(as.integer(unexpectedValues)),
            firstMicrobe,
            lastMicrobe)
    reportVec["UnexpectedValuesInED"] <- TRUE
  }

  return(reportVec)
}


#[END]