data/popproj.R

In wpp2015: World Population Prospects 2015

# This dataset is created on the fly as a sum of the age-specific population median projections popMprojMed and popFprojMed

popproj <- local({
	source('popMprojMed.R')
	source('popFprojMed.R')
	
	sum.by.country <- function(dataset) {
		year.cols.idx <- grep('^[0-9]{4}', colnames(dataset))
		plyr::ddply(dataset[,c(which(colnames(dataset)=='country_code'), year.cols.idx)], 
		            "country_code", .fun = plyr::colwise(sum))
	}
	tpopM <- sum.by.country(popMprojMed)
	tpopF <- sum.by.country(popFprojMed)
	# The male and female dataset should be in the same format, 
	# i.e. the countries and years should be in the same order, but just to be sure
	# match columns and rows. It will fail if there are different sets of countries
	# in the two datasets.
	cols.to.sumM <- colnames(tpopM)[-match('country_code', colnames(tpopM))]
	cols.to.sumF <- colnames(tpopF)[-match('country_code', colnames(tpopF))]
	cols.to.sumF.idx <- match(cols.to.sumF, cols.to.sumM)
	rowsF.idx <- match(tpopF$country_code, tpopM$country_code)
	name.col <- grep('^name$|^country$', colnames(popMprojMed), value=TRUE)
	cbind(country_code=tpopM$country_code, 
				 name=popMprojMed[,name.col][match(tpopM$country_code, popMprojMed$country_code)],
				 tpopM[,cols.to.sumM] + tpopF[rowsF.idx, cols.to.sumF[cols.to.sumF.idx]])
})