# align data extracted by Had

#library(rgdal)
library(terra)
library(sf)

# Layers
# 1. Cassava production
# 2. Harvest Area
# 3. ls population 2014
# 4. settlement 2018
# extracted at 2km, 5km and 10km buffers

# Point buffers.
# original outputs had each settlement layer stacked with labels in betwen. Manually manipulated these to be in wide format

CIV_buff_pointstats_max <- read.csv("CIV_buff_pointstats_wide_max.csv", header=TRUE, skip=2)
CIV_buff_pointstats_min <- read.csv("CIV_buff_pointstats_wide_min.csv", header=TRUE, skip=2)
CIV_buff_pointstats_mean <- read.csv("CIV_buff_pointstats_wide_mean.csv", header=TRUE, skip=2)
CIV_buff_pointstats_sd <- read.csv("CIV_buff_pointstats_wide_sd.csv", header=TRUE, skip=2)
CIV_buff_pointstats_sum <- read.csv("CIV_buff_pointstats_wide_sum.csv", header=TRUE, skip=2)
CIV_buff_pointstats_quantile <- read.csv("CIV_buff_pointstats_wide_quantile.csv", header=TRUE, skip=2)

UGA_buff_pointstats_max <- read.csv("UGA_buff_pointstats_wide_max.csv", header=TRUE, skip=2)
UGA_buff_pointstats_min <- read.csv("UGA_buff_pointstats_wide_min.csv", header=TRUE, skip=2)
UGA_buff_pointstats_mean <- read.csv("UGA_buff_pointstats_wide_mean.csv", header=TRUE, skip=2)
UGA_buff_pointstats_sd <- read.csv("UGA_buff_pointstats_wide_sd.csv", header=TRUE, skip=2)
UGA_buff_pointstats_sum <- read.csv("UGA_buff_pointstats_wide_sum.csv", header=TRUE, skip=2)
UGA_buff_pointstats_quantile <- read.csv("UGA_buff_pointstats_wide_quantile.csv", header=TRUE, skip=2)


# tidy these up, by removing multiple sample ids and merging together per country

CIV_buff_pointstats_max[, seq(from=3, by=2, to=23)] <- NULL
CIV_buff_pointstats_min[, seq(from=3, by=2, to=23)] <- NULL
CIV_buff_pointstats_mean[, seq(from=3, by=2, to=23)] <- NULL
CIV_buff_pointstats_sd[, seq(from=3, by=2, to=23)] <- NULL
CIV_buff_pointstats_sum[, seq(from=3, by=2, to=23)] <- NULL
CIV_buff_pointstats_quantile[, seq(from=7, by=6, to=72)] <- NULL

names(CIV_buff_pointstats_max) <- c("SampleID", paste("max_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(CIV_buff_pointstats_min) <- c("SampleID", paste("min_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(CIV_buff_pointstats_mean) <- c("SampleID", paste("mean_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(CIV_buff_pointstats_sd) <- c("SampleID", paste("sd_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(CIV_buff_pointstats_sum) <- c("SampleID", paste("sum_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(CIV_buff_pointstats_quantile) <- c("SampleID", paste(rep(c("min", "lowerQ", "median", "upperQ", "max"), times=4*3), rep(c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), each=5), sep="_"))

CIV_buff_pointstats <- merge(CIV_buff_pointstats_mean, CIV_buff_pointstats_sd, by="SampleID")
CIV_buff_pointstats <- merge(CIV_buff_pointstats, CIV_buff_pointstats_sum, by="SampleID")
CIV_buff_pointstats <- merge(CIV_buff_pointstats, CIV_buff_pointstats_quantile, by="SampleID")

dim(CIV_buff_pointstats)




UGA_buff_pointstats_max[, seq(from=3, by=2, to=23)] <- NULL
UGA_buff_pointstats_min[, seq(from=3, by=2, to=23)] <- NULL
UGA_buff_pointstats_mean[, seq(from=3, by=2, to=23)] <- NULL
UGA_buff_pointstats_sd[, seq(from=3, by=2, to=23)] <- NULL
UGA_buff_pointstats_sum[, seq(from=3, by=2, to=23)] <- NULL
UGA_buff_pointstats_quantile[, seq(from=7, by=6, to=72)] <- NULL

names(UGA_buff_pointstats_max) <- c("SampleID", paste("max_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(UGA_buff_pointstats_min) <- c("SampleID", paste("min_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(UGA_buff_pointstats_mean) <- c("SampleID", paste("mean_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(UGA_buff_pointstats_sd) <- c("SampleID", paste("sd_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(UGA_buff_pointstats_sum) <- c("SampleID", paste("sum_", c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), sep=""))
names(UGA_buff_pointstats_quantile) <- c("SampleID", paste(rep(c("min", "lowerQ", "median", "upperQ", "max"), times=4*3), rep(c("Prod.2000", "Prod.5000", "Prod.10000", "Harv.2000", "Harv.5000", "Harv.10000", "lspop.2000", "lspop.5000", "lspop.10000", "settle.2000", "settle.5000", "settle.10000"), each=5), sep="_"))

UGA_buff_pointstats <- merge(UGA_buff_pointstats_mean, UGA_buff_pointstats_sd, by="SampleID")
UGA_buff_pointstats <- merge(UGA_buff_pointstats, UGA_buff_pointstats_sum, by="SampleID")
UGA_buff_pointstats <- merge(UGA_buff_pointstats, UGA_buff_pointstats_quantile, by="SampleID")

dim(UGA_buff_pointstats)




# check long and lat of the sample points
CIV_pnts = st_read(dsn="cassava_wip.gdb",layer=paste0("CIV","_centroids"))
CIV_pnts$SampleID <- 1:nrow(CIV_pnts)
CIV_buff_pointstats <- merge(CIV_buff_pointstats, CIV_pnts[, c("SampleID", "Centre_Lon", "Centre_Lat")], by="SampleID")


UGA_pnts = st_read(dsn="cassava_wip.gdb",layer=paste0("UGA","_centroids"))
UGA_pnts$SampleID <- 1:nrow(UGA_pnts)
UGA_buff_pointstats <- merge(UGA_buff_pointstats, UGA_pnts[, c("SampleID", "Centre_Lon", "Centre_Lat")], by="SampleID")


rm(list=ls()[which(!(ls()%in%c("CIV_buff_pointstats","UGA_buff_pointstats","civ","uga","dat_civ","dat_uga")))])