R/correctIndiaAPY.R

In pik-piam/mrfable: FABLE project data

#' @title correctIndiaAPY
#' @description Correct IndiaAPY data
#' @param x magpie object provided by the read function
#' @return magpie objects with corrected IndiaAPY data
#' @author Anastasis Giannousakis
#' @examples
#' \dontrun{
#' readSource("IndiaAPY", convert = "onlycorrect")
#' }
#'
#' @importFrom magclass as.magpie as.data.frame getCells getCells<- dimSums mbind getNames<-
#' @importFrom utils read.csv
#' @importFrom madrat toolCountryFill
#' @importFrom dplyr  %>% select anti_join filter mutate ungroup group_by summarise bind_rows distinct

correctIndiaAPY <- function(x) {

  .rmvdplseas <- function(xx, seas1, seas2) {
    tmp <- select(filter(xx, season %in% c(seas1, seas2)), -"season")
    tmp <- tmp[which(duplicated(tmp)), ]
    if (nrow(tmp) > 0) xx <- anti_join(xx, mutate(tmp, "season" = seas2),
        by = c("state", "year", "crop", "variable", "unit", "season", "Value"))
    return(xx)
  }

  x <- as.data.frame(x) %>% select(-"Cell")
  x <- x[which(!is.na(x[, "Value"])), ]

  x[, "Region"] <- sub("Chattisgarh|Chhatisgarh", "Chhattisgarh", x[, "Region"])
  x[, "Region"] <- sub("UttaraKhand", "Uttarakhand", x[, "Region"])
  x[, "Region"] <- sub("Dadra Nagar Haveli", "D & N Haveli", x[, "Region"])
  x[, "Region"] <- sub("Panjab", "Punjab", x[, "Region"])
  x[, "Region"] <- sub("Pondicherry", "Puducherry", x[, "Region"])
  x[, "Region"] <- sub("Orissa", "Odisha", x[, "Region"])
  x[, "Region"] <- sub("A & N Islands", "Andaman and Nicobar Islands", x[, "Region"])
  x[, "Data4"] <- sub("NA", "total", x[, "Data4"])
  x[, "Data4"] <- sub("total kharif", "kharif total", x[, "Data4"])
  x[, "Data4"] <- sub("/^kharif/", "kharif total", x[, "Data4"])
  x[, "Data4"] <- sub("kharif \\(autumn\\)", "kharif total", x[, "Data4"])
  x[, "Data4"] <- sub("kharif \\(winter\\)", "kharif total", x[, "Data4"])
  x[, "Data4"] <- sub("summer/rabi", "rabi/summer", x[, "Data4"])

  colnames(x) <- c("state", "year", "crop", "variable", "unit", "season", "Value")

  season <- state <- variable <- crop <- year <- Value <- unit <- NULL #nolint
  x <- filter(x, season != "total rabi/ summer")

  # remove mistakes in excel files leading to double counting (e.g. Rice,2015,Jharkhand)
  x <- .rmvdplseas(x, "autumn", "kharif")
  x <- .rmvdplseas(x, "winter", "kharif")

  # add "kharif total" only where it is missing.
  # (because some regions/years/crops have it as a new data point
  # whereas most of them calculate it in the excel files)
  x %>%
    filter(season %in% c("kharif", "autumn", "winter")) %>%
    group_by(`season`, `state`, `crop`, `year`, `variable`, `unit`) %>%
    summarise("Value" = sum(Value, na.rm = TRUE)) %>%
    mutate(season = "kharif total") %>%
    ungroup() -> tmp
  tmp1 <- bind_rows(filter(x, season == "kharif total"), tmp)
  tmp1 %>% distinct() %>%  distinct(`state`, `year`, `crop`, `unit`, `season`, .keep_all = T) -> tmp
  x <- bind_rows(tmp, filter(x, season != "kharif total"))

  # add "total" where it is missing
  x %>%
    filter(!season %in% c("kharif total", "total")) %>%
    group_by(`season`, `state`, `crop`, `year`, `variable`, `unit`) %>%
    summarise("Value" = sum(`Value`, na.rm = TRUE)) %>%
    mutate(season = "total") %>%
    ungroup() -> tmp
  tmp1 <- bind_rows(filter(x, season == "total"), tmp)
  tmp1 %>% distinct() %>%  distinct(`state`, `year`, `crop`, `unit`, `season`, .keep_all = T) -> tmp
  x <- bind_rows(tmp, filter(x, season != "total"))

  # add "total" if there is only "kharif total" as season
  x %>% filter(season %in% c("kharif total")) -> tmp
  tmp[, "season"] <- "total"
  tmp1 <- bind_rows(x, tmp)
  tmp1 %>% distinct() %>%  distinct(`state`, `year`, `crop`, `unit`, `season`, .keep_all = T) -> x

  x <- as.magpie(x, spatial = "state")

  x["D & N Haveli", , ] <- dimSums(x[c("D & N Haveli", "Daman & Diu"), , ], dim = 1, na.rm = TRUE)

  getCells(x) <- sub("D & N Haveli", "Dadra and Nagar Haveli and Daman and Diu", getCells(x))
  mapping <- read.csv(system.file("extdata", "regional/mappingIndiaAPY.csv", package = "mrfable"))
  x <- toolCountryFill(x, countrylist = as.vector(mapping[, "state"]),
                       no_remove_warning = c("All India", "Daman & Diu"))

  getNames(x) <- sub("kHektares", "kHectares", getNames(x))

  return(x)
}