R/convertIFA.R

In pik-piam/mrcommons: MadRat commons Input Data Library

#' Convert IFA
#'
#' Convert IFADavies (2013) data to ISO country level.
#'
#'
#' @param x MAgPIE object containing IFA data region resolution
#' @return MAgPIE object of the IFA data disaggregated to country level
#' @author Lavinia Baumstark
#' @examples
#' \dontrun{
#' a <- convertIFA(x)
#' }
#'
#' @importFrom magclass where
#' @importFrom magpiesets findset

convertIFA <- function(x) {
  # delete all non-grand-totals
  x <- x[, , c("Grand Total N", "Grand Total P2O5", "Grand Total K2O")]

  # delete dots in names as they cause problems with dimensions
  getNames(x) <- gsub("\\.", "", getNames(x))

  mapping <- toolGetMapping(type = "regional", name = "transition_mapping_IFA.csv", where = "mappingfolder")

  # add 0 for azerbaijan in 1990,
  vcat(verbosity = 1, "azerbaijan is set to 0 in 1990 to allow for toolISOhistorical")
  x["AZE", , "Grand Total P2O5"] <- 0
  x["SRB", 1994:1995, "Grand Total K2O"] <- setYears(x["SRB", 1996, "Grand Total K2O"], NULL)
  x["BIH", 1994, ] <- setYears(x["BIH", 1996, ], NULL)
  x["SVN", 1993, ] <- NA

  # convert historical data to countries with ISO-codes
  x <- toolISOhistorical(x, mapping = mapping)

  # disaggregate the Others* regions of the data
  mapping <- "regionmappingIFA.csv"
  m <- read.csv2(mapping)
  weight <- calcOutput("Population", aggregate = FALSE)[as.character(m$CountryCode), 2010, "pop_SSP2"]
  xOthers <- toolAggregate(x[as.character(unique(m$RegionCode)), , ], mapping, weight = weight)
  # add countries of Others*-Regions
  out <- mbind(x, xOthers)

  out <- toolCountryFill(out, fill = 0, no_remove_warning = c("XOA", "XEA", "XOO", "XWA", "XAC"))

  vcat(verbosity = 2, paste("Incomplete fertilizer data in", paste(where(is.na(out))$true$regions, collapse = " ")))
  vcat(
    verbosity = 1,
    paste(
      "Incomplete fertilizer data in 5year timesteps in",
      paste(where(is.na(out[, findset("past"), ]))$true$regions, collapse = " ")
    )
  )
  vcat(
    verbosity = 1,
    paste(
      "Incomplete fertilizer data in 5year timesteps in 2010 in ",
      paste(where(is.na(out[, "y2010", ]))$true$regions, collapse = " "),
      ", filled with 0s."
    )
  )

  out[is.na(out)] <- 0
  out <- out / 1000

  return(out)
}