R/Bootstrap_Moran_e_MESS_TMLA.R

In andrefaa/ENM_TheMetaLand: Create and process Ecological Niche, including several pre- and post-processing methods

### This is the R code to conduct ENMs evaluation by the Odds-and-Evens Framework
# In the moment, the code is configured to create the splits only by its latitudinal axis
# Read carefully the Initialization and function parameters and the messages that appear on the console
# Developers: Paulo De Marco Junior & Andre Andrade

Bootstrap_Moran_e_MESS_TMLA <- function(Env,
                                        RecordsData,
                                        DirO,
                                        repl) {
  #Function Parameters:
  #env:environmental variables stack
  #occ: occurrence data

  #Important Observations:
  #Occurence File MUST have the columns named Species (containing species name),Long (containing longitude),Lat (containing latitude)

  #Development
  occTR <- RecordsData[RecordsData$PresAbse == 1, ]

  # Lists for validation-----
  assign(paste("Bootstrap_Moran_MESS", sep = ""), list())
  VALNAME <- paste0("Bootstrap_Moran_MESS_", repl, ".txt")

  #Calculate Moran's I
  Moran <- Moran_for_Bootstrap_TMLA(occTR = occTR, pc1 = Env[[1]])

  #MESS----
  Occ_e <- cbind(occTR, raster::extract(Env, occTR[, 2:3]))
  Occ_e <- split(Occ_e, f = Occ_e$sp)
  Occ_e <- lapply(Occ_e, function(x)
    split(x, f = x$Partition))

  mess <-
    lapply(Occ_e, function(x)
      MESS(x[[1]][, -c(1:5)], x[[2]][, -c(1:5)]))
  mess <-
    ldply(lapply(mess, function(x)
      mean(x$TOTAL, na.rm = TRUE)), .id = NULL)
  Bootstrap_Moran_MESS[[1]] <- data.frame(Moran, MESS = mess)

  # Save .txt with models Moran&MESS----
  Obj <- ls(pattern = 'Bootstrap_Moran_MESS')
  res <- list()
  for (i in 1:length(Obj)) {
    res[[i]] <- plyr::ldply(get(Obj[i]))
  }
  res <- plyr::ldply(res)
  colnames(res)[3] <- "MESS"
  if (is.null(repl) == F) {
    res <- data.frame(res, Replicate = repl)
  }
  utils::write.table(
    res,
    file.path(DirO, VALNAME),
    sep = "\t",
    row.names = F,
    col.names = T
  )
}