Script to reformat water level predictions from Thai Navy monitoring stations

reformData.R

### script to reformat water level predictions from Thai Navy monitoring stations

reformatData <- function(filename, timezone = "GMT") {
  ### filename = location of tab-delimited text file with Thai Navy water level predictions
  ### timezone = I'm not sure how to refer to the local time zone, so I used GMT to at least avoid daylight savings time issues
  
  ### usage example
  # UdayDat <- "C:/RDATA/Thailand/data/data_waterLevel2018.txt"
  # dat1 <- reformatData(filename = UdayDat)
  # tail(dat1)

  ### check for reshape2, install if not available
  list.of.packages <- c("reshape2")
  new.packages <- list.of.packages[!(list.of.packages %in% installed.packages()[,"Package"])]
  if(length(new.packages)) install.packages(new.packages)
  

  wtr <- read.delim(filename, stringsAsFactors = FALSE)
  
  ### identify rows where months begin using "DATE" located above days
  month.start       <- c(grep(pattern = "DATE", x = wtr$X), nrow(wtr) + 8 )
  month.name.col    <-  wtr$X121
  months.in.dataset <- month.name.col[grep(x = month.name.col, pattern = paste(month.name, collapse = "|"))]

  ### process month by month
  for (j in 1:length(months.in.dataset)) {
    monthName <- months.in.dataset[j]
    sub1 <- wtr[(month.start[j] + 2):(month.start[j+1] - 8), c(1, 3:26)]
    sub1 <- sub1[complete.cases(sub1), ]
    
    names(sub1) <- c("day", paste0("h", 0:23))
    
    for (i in 1:ncol(sub1)) {
      sub1[, i] <- as.numeric(as.character(sub1[, i]))
    }
    
    sub2          <- reshape2::melt(data = sub1, id.vars = "day", value.name = "meters.MLLW")
    sub2$hour     <- as.numeric(substr(sub2$variable, 2, 3))
    sub2$datetime <- as.POSIXct(paste(sep = " ", monthName, sub2$day, sub2$hour), format = "%B %Y %d %H", tz = timezone)
    
    finSub <- sub2[, c("datetime", "meters.MLLW")]
    
    if (j == 1) {
      outDat <- finSub
    } else {
      outDat <- rbind(outDat, finSub)
    }
  }
  
  outDat <- outDat[order(outDat$datetime), ]
  row.names(outDat) <- 1:nrow(outDat)
  
  invisible(outDat)
}