require(data.table)

cropIntensitySummary = function (luDt, cropType) {
	
	yDt = luDt[cropland > 0, 
			list(	irrig = get(paste0(cropType, '_IQ')), 
					fert = get(paste0(cropType, '_FQ')),
					otherint = get(paste0(cropType, '_OI')),
					thisCropFact = get(paste0(cropType, '_A')), 
					allCropFact = cropland/area, 
					area = get(paste0(cropType, '_A'))*cropland, 
					yield = get(paste0(cropType, '_Y'))) ]
	
#	yDt[thisCropFact>0.05 & allCropFact > 0.03, list(cropType = cropType, yield=sum(area * yield)/sum(area), 
	yDt[thisCropFact>0, list(cropType = cropType, yield=sum(area * yield)/sum(area), 
					irrig=sum(area * irrig)/sum(area), 
					fert=sum(area * fert)/sum(area),
					otherint=sum(area * otherint)/sum(area),
					area = sum(area)) ]
}

summariseLandUseOutput = function(simDir, writeFile=TRUE) {
	
	years = list.files(simDir, pattern="[0-9]+")
	cropSumDt = NULL
	
	for (year in years) {
		fileName = file.path(simDir, year, 'LandUse.txt')
		if (file.exists(fileName)) {
			print (paste("Processing", year, "from", fileName))
			luDt = data.table(read.table(fileName, header=TRUE))
		
			for (cropType in c('maize', 'rice', 'wheat', 'energycrops', 'oilcrops', 'pulses', 'starchyRoots')) {
				cropSummary = cropIntensitySummary(luDt, cropType)
				if (nrow(cropSummary) >0)
					cropSumDt = rbind(cropSumDt, cbind(cropSummary, year=year))
			}
		} else {
			print(paste("Can't find file, so skipping:", fileName))
		}
	}

	resDt = rbind(cropSumDt, cropSumDt[cropType != 'energycrops', list(cropType='avgExcEC', 
					yield=sum(yield*area)/sum(area), 
					irrig=sum(irrig*area)/sum(area),
					fert=sum(fert*area)/sum(area),
					otherint=sum(otherint*area)/sum(area),
					area=sum(area)), by=year])

	resDt = rbind(resDt, cropSumDt[, list(cropType='avgIncEC', 
						yield=sum(yield*area)/sum(area), 
						irrig=sum(irrig*area)/sum(area),
						fert=sum(fert*area)/sum(area),
						otherint=sum(otherint*area)/sum(area),
						area=sum(area)), by=year])

	if (writeFile) write.table(resDt, file.path(simDir, "LandUseSummary.csv"), sep=",", row.names=FALSE, quote=FALSE)
	resDt
}

baseOutputDir="/exports/csce/eddie/geos/groups/LURG/models/PLUM/output" # "~/Downloads"

ensemble = commandArgs(trailingOnly = TRUE)[1]
ensDir=file.path(baseOutputDir, ensemble)
if (dir.exists(ensDir)) {
	print(paste(ensDir, "exists.  Processing"))
	summariseLandUseOutput(simDir = ensDir)
} else {
	print(paste(ensDir, "does not exist.  Stopping"))
}