data-raw/02_generating_maps.R
In LimaRAF/plantR: Managing Species Records from Biological Collections
#### Script to read shapefiles, edit and generate the data/worldMap.rds ####
require(rgdal)
require(rgeos)
require(sf)
require(sp)
require(countrycode)
devtools::load_all()
##List of Neotropical countries
countries.latam <- c("Anguilla","Antigua and Barbuda","Argentina","Aruba","Bahamas",
"Barbados","Belize","Bermuda","Bolivia","Brazil","British Virgin Islands",
"Caribbean Netherlands","Cayman Islands","Chile",
"Colombia","Costa Rica","Cuba","Curaçao","Dominica","Dominican Republic",
"Ecuador","El Salvador","Falkland Islands",
"French Guiana","Grenada","Guadeloupe","Guatemala","Guyana","Haiti",
"Honduras","Jamaica","Martinique","Mexico","Montserrat","Nicaragua",
"Panama","Paraguay","Peru","Puerto Rico","Saint Barthélemy",
"Saint Kitts and Nevis","Saint Lucia","Saint Martin (French part)","Saint Vincent and the Grenadines",
"Sint Maarten","Suriname","Trinidad and Tobago","Turks and Caicos Islands",
"United States Virgin Islands","Uruguay","Venezuela")
iso3 <- countrycode::countrycode(countries.latam, "country.name", "iso3c")
#BES = "Caribbean Netherlands" = "Bonaire, Saint Eustatius and Saba"
################################################################################H
################################################################################H
################################
### WORLD MAP FOR VALIDATION ###
################################
##new sf files
destfolder <- "vrac/latam"
destfolder <- "data-raw/GDAM"
if (!exists(destfolder)) dir.create(destfolder, recursive = T)
source("https://raw.githubusercontent.com/liibre/Rocc/master/R/getGADM.R")
#Dando problema
# purrr::walk2(.x = rep(iso3, each = 4),
# .y = rep(3:0, length(countries.latam)),
# ~ getGADM(
# cod = .x,
# level = .y,
# best = FALSE, #importante para ter outros niveis
# destfolder = destfolder
# ))
gadm_files <- list.files(destfolder, pattern = "rds$",
full.names = TRUE)
##Loading, editing and converting the world country shapefile
path <- "E://ownCloud//W_GIS"
#wo <- readOGR(dsn=paste(path,"//WO_ADM_limits_GADM36//gadm36_levels_shp",sep=""),layer="gadm36_0")
wo <- readRDS(paste(path,"//WO_ADM_limits_GADM36//gadm36_levels_shp//gadm36_0.rds",sep = ""))
#wo <- readRDS(paste0(path,"//WO_ADM_limits_GADM36//gadm36_levels_shp//gadm36_0.rds"))
#Editing country name as in the gazetteer
wo@data$pais <- countrycode(as.character(wo@data$GID_0),'iso3c', 'country.name')
wo@data$pais[is.na(wo@data$pais)] <- as.character(wo@data$NAME_0[is.na(wo@data$pais)])
wo@data$pais <- prepCountry(wo@data$pais)
# wo@data$pais <- tolower(chartr(paste(names(unwanted_array), collapse=''), paste(unwanted_array, collapse=''), wo@data$pais))
# wo@data$pais <- tolower(textclean::replace_non_ascii(wo@data$pais))
# wo@data$pais <- stringr::str_replace_all(wo@data$pais, " & ", " and ")
# wo@data$pais <- gsub("^st. ", "saint ", wo@data$pais)
# wo@data$pais <- gsub(" of the ", " ", wo@data$pais)
# wo@data$pais <- gsub(" of ", " ", wo@data$pais)
tmp1 <- replaceNames[replaceNames$class %in% "country" & apply(is.na(replaceNames[, 2:4]), 1, all), ]
tmp2 <- as.character(tmp1$replace)
names(tmp2) = as.character(tmp1$pattern)
#names(tmp2) <- gsub("\\\\", "", names(tmp2))
wo@data$pais <- sapply(strsplit(wo@data$pais,' \\('), function(x) x[[1]][1])
wo@data$pais <- stringr::str_replace_all(wo@data$pais, tmp2)
# wo@data$pais <- plantR::prepLoc(wo@data$pais)
## Comparing the map with the gazetteer
tmp <- merge(wo@data, gazetteer[,c("loc", "loc.correct")], by.x = "pais", by.y = "loc", all.x = TRUE)
tmp[!tmp$pais %in% tmp$loc.correct,] #ok!
#Transforming and simplifying
wo@data <- wo@data[,c("GID_0","pais")]
names(wo@data)[2] <- "NAME_0"
wo.simp <- gSimplify(wo, tol=0.001, topologyPreserve = TRUE)
wo.simp <- gBuffer(wo.simp, byid = TRUE, width = 0)
wo.simp1 <- SpatialPolygonsDataFrame(wo.simp, wo@data)
wo.simp1_sf <- sf::st_as_sf(wo.simp1)
# set spatial coordinates
prj <- sf::st_crs(4326)
worldMap <- sf::st_set_crs(wo.simp1_sf, prj)
# wo.simp2 <- gSimplify(wo, tol = 0.0001, topologyPreserve = TRUE)
# wo.simp2 <- gBuffer(wo.simp2, byid = TRUE, width = 0)
# wo.simp3 <- SpatialPolygonsDataFrame(wo.simp2, wo@data)
# worldMap_0001 <- sf::st_as_sf(wo.simp3)
# wo.simp4 <- gSimplify(wo, tol = 0.1, topologyPreserve = TRUE)
# wo.simp4 <- gBuffer(wo.simp4, byid = TRUE, width = 0)
# wo.simp5 <- SpatialPolygonsDataFrame(wo.simp4, wo@data)
# worldMap_01 <- sf::st_as_sf(wo.simp5)
# #wo.simp2 <- sf::st_buffer(wo.simp1, dist = 0)
# worldMap_full <- sf::st_as_sf(wo)
#
# #Inspecting
# par(mfrow = c(2,2), mar=c(0,0,1,0))
# plot(wo[1,], main="Aruba original")
# plot(wo.simp[1,], border = "red", main="Aruba gSimp 0.001")
# plot(st_geometry(worldMap[1,1]), border = "green", main = "Aruba gSimp 0.001 sf")
#Repairing possible issues related to spherical geometries
isv <- sf::st_is_valid(worldMap) #Senegal had an issue
if (any(isv)) {
for (i in which(!isv)) {
tmp.i <- worldMap[i, ]$geometry
tmp.i.1 <- s2::as_s2_geography(sf::st_as_binary(tmp.i), check = FALSE)
tmp.i.2 <- s2::s2_as_binary(s2::s2_snap_to_grid(tmp.i.1, grid_size = 1e-7))
tmp.i.3 <- sf::st_as_sfc(s2::s2_geog_from_wkb(tmp.i.2, check = TRUE))
worldMap[i, ]$geometry <- tmp.i.3
}
}
# worldMap <- sf::st_make_valid(worldMap)
#Saving
save(worldMap, file = "./data/worldMap.rda", compress = "xz")
# save(worldMap_0001, file = "./data/worldMap_0001.rda", compress = "xz")
# save(worldMap_full, file = "./data/worldMap_full.rda", compress = "xz")
################################################################################H
################################################################################H
#########################################
### LATIN AMERICAN MAP FOR VALIDATION ###
#########################################
##Loading, editing and converting the world country shapefile
path0 <- "E://ownCloud//W_GIS//Am_Lat_ADM_GADM_v3.6//"
countries <- list.files(path0, full.name=TRUE)
countries <- countries[grep(paste(iso3, collapse = "|") ,countries)]
countries <- countries[grep("sf.rds",countries)]
countries3 <- countries[grep('_3_sf.rds',countries)]
adm3 <- sapply(strsplit(countries3, "//|_"), function(x) x[11])
countries2 <- countries[grepl('_2_sf.rds',countries) & !grepl(paste0(c(adm3), collapse = "|"), countries)]
adm2 <- sapply(strsplit(countries2, "//|_"), function(x) x[11])
countries1 <- countries[grepl('_1_sf.rds',countries) & !grepl(paste0(c(adm2,adm3), collapse = "|"), countries)]
adm1 <- sapply(strsplit(countries1, "//|_"), function(x) x[11])
countries0 <- countries[grepl('_0_sf.rds',countries) & !grepl(paste0(c(adm1,adm2,adm3), collapse = "|"), countries)]
adm0 <- sapply(strsplit(countries0, "//|_"), function(x) x[11])
all.countries <- sort(c(adm1, adm2, adm3))
iso3[!iso3 %in% all.countries] # all of these countries are only available at ADM_0, that is, already in worldMap.
countries <- sort(c(countries1, countries2, countries3))
country.list <- vector('list', length(countries))
pais = NULL
# wo <- readRDS(gadm_files)
# wo <- purrr::map(gadm_files, .f = function(x) readRDS(file = x))
for (i in 1:length(country.list)) {
path = countries[i]
tmp = readRDS(path)
#tmp1 = tmp@data
tmp1 = tmp[,grepl("^NAME_", names(tmp))]
#country
tmp1$NAME_0 <- prepCountry(tmp1$NAME_0)
# tmp1$NAME_0 <- tolower(textclean::replace_non_ascii(tmp1$NAME_0))
# tmp1$NAME_0 <- stringr::str_replace_all(tmp1$NAME_0, " & ", " and ")
# tmp1$NAME_0 <- gsub("^st. ", "saint ", tmp1$NAME_0)
# tmp1$NAME_0 <- gsub(" of the ", " ", tmp1$NAME_0)
# tmp1$NAME_0 <- gsub(" of ", " ", tmp1$NAME_0)
toto1 <- replaceNames[replaceNames$class %in% "country" & apply(is.na(replaceNames[, 2:4]), 1, all), ]
toto2 <- as.character(toto1$replace)
names(toto2) = as.character(toto1$pattern)
# names(toto2) <- gsub("\\\\", "", names(toto2))
tmp1$NAME_0 <- sapply(strsplit(tmp1$NAME_0,' \\('), function(x) x[[1]][1])
tmp1$NAME_0 <- stringr::str_replace_all(tmp1$NAME_0, toto2)
# tmp1$NAME_0 <- plantR::prepLoc(tmp1$NAME_0)
tmp1$NAME_0 <- gsub("bonaire, sint eustatius and saba", "caribbean netherlands", tmp1$NAME_0)
tmp1$NAME_0 <- gsub("bonaire, sint eustatius saba", "caribbean netherlands", tmp1$NAME_0)
tmp1$NAME_0 <- gsub("saint vincent and the grenadines", "saint vincent grenadines", tmp1$NAME_0)
tmp1$NAME_0 <- gsub("virgin islands, u.s.", "united states virgin islands", tmp1$NAME_0)
#ADM_1
tmp1$NAME_1 <- tolower(textclean::replace_non_ascii(tmp1$NAME_1))
tmp1$NAME_1 <- plantR::prepLoc(tmp1$NAME_1)
if ("NAME_2" %in% names(tmp1)) {
tmp1$NAME_2 <- tolower(textclean::replace_non_ascii(tmp1$NAME_2))
tmp1$NAME_2 <- plantR::prepLoc(tmp1$NAME_2)
}
if ("NAME_3" %in% names(tmp1)) {
tmp1$NAME_3 <- tolower(textclean::replace_non_ascii(tmp1$NAME_3))
tmp1$NAME_3 <- plantR::prepLoc(tmp1$NAME_3)
}
#Symplifying the maps
# tmp.simp <- gSimplify(tmp, tol=0.0001, topologyPreserve = TRUE)
# tmp.simp <- gBuffer(tmp.simp, byid=TRUE, width=0)
tmp.simp <- st_simplify(tmp1, dTolerance = 0.0001, preserveTopology = TRUE)
tmp.simp <- st_buffer(tmp.simp, dist=0)
# tmp.simp <- tmp1
##Simplify transform MULTIPOLYGONS in POLYGON: is that a problem?
#Extract polygon ID's
# pid <- sapply(slot(tmp.simp, "polygons"), function(x) slot(x, "ID"))
#
# #Create dataframe with correct rownames and merging
# p.df <- as.data.frame(tmp1, row.names = pid, stringsAsFactors = FALSE)
# tmp.simp1 <- SpatialPolygonsDataFrame(tmp.simp, p.df)
# set spatial coordinates
prj <- sf::st_crs(4326)
tmp.simp <- sf::st_set_crs(tmp.simp, prj)
#Saving
# pais[i] = unique(p.df$NAME_0)
pais[i] = unique(tmp.simp$NAME_0)
country.list[[i]] = tmp.simp
}
names(country.list) = pais
#pais[!pais %in% wo.simp1$NAME_0]
## Standardizing the shapefile and gazetteer names (for Brazil only)
# #loading the gazetteer and removing possible duplicated localities
# dic <- read.csv("data-raw//raw_dictionaries//gazetteer.csv",as.is=TRUE)
# # dic <- read.csv("C://Users//renato//Documents//raflima//Pos Doc//Manuscritos//Artigo AF checklist//data analysis//dictionaries//gazetteer.csv",as.is=TRUE)
# dic <- dic[dic$status %in% "ok",]
# dic <- dic[dic$resolution.gazetteer %in% c("country","state","county"),]
#
# j = which(names(country.list) == "brazil")
# #for(j in 31:length(country.list)) {
# #Creating the locality strings for the shapefile dataframe
# tmp = country.list[[j]]#@data
# #for(i in 1:dim(tmp)[2]) tmp[,i] <- as.character(tmp[,i])
#
# tmp.df <- tmp
# st_geometry(tmp.df) <- NULL
# # tmp1= data.frame(loc=apply(tmp, 1, paste, collapse="_"))
# ids <- names(tmp.df) %in% c("NAME_0","NAME_1","NAME_2")
# tmp1= data.frame(loc=apply(tmp.df[, ids], 1, paste, collapse="_"),
# stringsAsFactors = FALSE)
# # tmp1$loc = as.character(tmp1$loc)
#
# tmp$order = 1:dim(tmp)[1]
# tmp1$order = 1:dim(tmp1)[1]
#
# #merging the two sources of info
# tmp2 <- merge(tmp1, dic[,c("NAME_0","NAME_1","NAME_2","loc","loc.correct")],
# by="loc", all.x=TRUE)
# tmp3 <- strsplit(tmp2$loc.correct,"_")
#
# nms <- c("country", "state", "county")
# for(w in 1:sapply(tmp3, length)[1]) {
# n <- rep(NA, dim(tmp2)[1])
# n <- sapply(tmp3, function(x) x[w])
# df <- data.frame(n, stringsAsFactors = FALSE); names(df) = nms[w]
# tmp2 <- cbind.data.frame(tmp2, df, stringsAsFactors = FALSE)
# }
#
# #comparing and replacing the correct info
# tmp2 <- tmp2[match(tmp$order, tmp2$order),]
# table(tmp$order == tmp2$order)
# id = !tmp2$loc %in% tmp2$loc.correct
#
# if(any(id)) {
# cat(paste(i,": ",unique(tmp$NAME_0),table(id)),"\n")
#
# table(tmp$NAME_0[id] == tmp2[id,c("country")])
# #tmp[id,1] = tmp2[id,c("country")]
# table(tmp$NAME_1[id] == tmp2[id,c("state")]) #all differences were double-checked and the gazetteer is right!
# cbind(tmp$NAME_1[id], tmp2[id,c("state","county")])[tmp$NAME_1[id] == tmp2[id,c("state")],]
# #tmp[id,2] = tmp2[id,c("state")]
# table(!tmp$NAME_2[id] == tmp2[id,c("county")]) #all differences were double-checked and the gazetteer is right!
#
# cbind(tmp$NAME_1[id],tmp$NAME_2[id], tmp2$county[id])[!tmp$NAME_2[id] == tmp2$county[id],]
# # tmp$NAME_2[id] = tmp2[id,c("county")]
#
# #Saving the changes
# country.list[[j]]$NAME_2[id] <- tmp2$county[id]
# # country.list[[j]]@data = tmp[,c("NAME_0","NAME_1","NAME_2")]
# }
# #}
# ## Other GADM issues and mixed names
# #Manaus
# tmp <- country.list[[j]][country.list[[j]]$NAME_2 %in% "maues",]
# tmp1 <- sf::st_coordinates(sf::st_centroid(tmp))
# ids.manaus <- tmp1[,1] < (-59) & tmp1[,2] > (-3.5)
# country.list[[j]]$NAME_2[country.list[[j]]$NAME_2 %in% "maues"][ids.manaus] <- "manaus"
#
# country.list[[j]]$check[country.list[[j]]$NAME_2 %in% "maues"] <- ids.manaus
#
# #Brasilândia de Minas/Brazópolis
# replace_this <- country.list[[j]]$NAME_2 %in% "brazopolis" &
# country.list[[j]]$NAME_3 %in% "brazilandia minas"
# country.list[[j]]$NAME_2[replace_this] <- "brasilandia minas"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #LEM
# replace_this <- country.list[[j]]$NAME_2 %in% "barreiras" &
# country.list[[j]]$NAME_3 %in% "luis eduardo magalhaes"
# country.list[[j]]$NAME_2[replace_this] <- "luis eduardo magalhaes"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Barrocas
# replace_this <- country.list[[j]]$NAME_2 %in% "serrinha" &
# country.list[[j]]$NAME_3 %in% "barrocas"
# country.list[[j]]$NAME_2[replace_this] <- "barrocas"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Capao Bonito do Sul
# replace_this <- country.list[[j]]$NAME_2 %in% "lagoa vermelha" &
# country.list[[j]]$NAME_3 %in% "capao bonito"
# country.list[[j]]$NAME_2[replace_this] <- "capao bonito sul"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Santa Cecília do Sul
# replace_this <- country.list[[j]]$NAME_2 %in% "tapejara" &
# country.list[[j]]$NAME_3 %in% "santa cecilia"
# country.list[[j]]$NAME_2[replace_this] <- "santa cecilia sul"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Balneario Rincão/SC
# replace_this <- country.list[[j]]$NAME_2 %in% "icara" &
# country.list[[j]]$NAME_3 %in% "balnerio rincao"
# country.list[[j]]$NAME_2[replace_this] <- "balneario rincao"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Três Barras/SC
# replace_this <- country.list[[j]]$NAME_2 %in% "sao mateus sul" &
# country.list[[j]]$NAME_3 %in% "tres barras"
# country.list[[j]]$NAME_1[replace_this] <- "santa catarina"
# country.list[[j]]$NAME_2[replace_this] <- "tres barras"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Teixeira Soares
# replace_this <- country.list[[j]]$NAME_2 %in% "texeira soares"
# country.list[[j]]$NAME_2[replace_this] <- "teixeira soares"
#
# #Mato Grosso as a county
# replace_this <- country.list[[j]]$NAME_2 %in% "mato grosso" &
# country.list[[j]]$NAME_3 %in% "aguapei"
# country.list[[j]]$NAME_2[replace_this] <- "vila bela santissima trindade"
#
# #Itabirinha and Itabirito/MG
# replace_this <- country.list[[j]]$NAME_2 %in% "itabirito" &
# country.list[[j]]$NAME_3 %in% "itabirinha mantena"
# country.list[[j]]$NAME_2[replace_this] <- "itabirinha"
#
#
# #Other counties
# map.checks <- readRDS("./data-raw/map.replacements.rds")
# paises <- sapply(map.checks$locs, function (x) strsplit(x, "_")[[1]][1])
# estados <- sapply(map.checks$locs, function (x) strsplit(x, "_")[[1]][2])
# j = which(names(country.list) == "brazil")
# for(w in 1:length(map.checks$locs)) {
# pais.i <- paises[w]
# estado.i <- estados[w]
# bad <- map.checks$municipio.bad[w]
# good <- map.checks$municipio.good[w]
# replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
# country.list[[j]]$NAME_1 %in% estado.i &
# country.list[[j]]$NAME_2 %in% bad &
# country.list[[j]]$NAME_3 %in% good
# country.list[[j]]$NAME_2[replace_this] <- good
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
# }
# Fixing GADM problems
map.checks <- read.csv("./data-raw/raw_dictionaries/gadmCheck.csv")
j = which(names(country.list) == "brazil")
for(w in 1:dim(map.checks)[1]) {
pais.i <- map.checks$pais[w]
estado.i <- map.checks$estado[w]
bad.i <- map.checks$bad.mun[w]
target.i <- map.checks$target.mun[w]
replace.i <- map.checks$replace.mun[w]
case.i <- map.checks$case[w]
if (case.i == "first") {
replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
country.list[[j]]$NAME_1 %in% estado.i &
country.list[[j]]$NAME_2 %in% bad.i &
country.list[[j]]$NAME_3 %in% target.i
if (any(replace_this))
country.list[[j]]$NAME_2[replace_this] <- replace.i
}
if (case.i == "second") {
replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
country.list[[j]]$NAME_1 %in% estado.i &
country.list[[j]]$NAME_2 %in% bad.i
if (any(replace_this))
country.list[[j]]$NAME_2[replace_this] <- replace.i
}
if (case.i == "county_state") {
replace.i.1 <- strsplit(replace.i,"\\|")[[1]][1]
replace.i.2 <- strsplit(replace.i,"\\|")[[1]][2]
replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
country.list[[j]]$NAME_1 %in% estado.i &
country.list[[j]]$NAME_2 %in% bad.i &
country.list[[j]]$NAME_3 %in% target.i
if (any(replace_this)) {
country.list[[j]]$NAME_1[replace_this] <- replace.i.1
country.list[[j]]$NAME_2[replace_this] <- replace.i.2
}
}
}
# ## Finding possible problems in GADM polygons: counties with disjoint polygons
# #Function that aggreagte polygons sharing their borders
# clusterSF <- function(sfpolys, thresh){
# df <- sfpolys[, grepl("NAME_", names(sfpolys))]
# sf::st_geometry(df) <- NULL
# nome <- unique(apply(df[,1:3], 1, paste, collapse = "_"))[1]
#
# dmat <- sf::st_distance(sfpolys)
# thresh.dist <- units::as_units('m', thresh)
# hc <- stats::hclust(as.dist(dmat > thresh.dist), method="single")
#
# groups = stats::cutree(hc, h=0.5)
# d = sf::st_sf(
# geom = do.call(c,
# lapply(1:max(groups), function(g){
# sf::st_union(sfpolys[groups==g,])
# })
# )
# )
# d$name <- nome
# d$group <- 1:nrow(d)
# d
# }
#
# #Selected country map
# country.pol <- country.list[[j]]
# # country.pol <- country.list[[which(names(country.list) == "brazil")]]
#
# #List of counties with more than one polygon
# toto <- paste(country.pol$NAME_1,country.pol$NAME_2, sep="_")
# possible.counties <- unique(toto[duplicated(toto)])
# #possible.counties <- possible.counties[country.pol$NAME_2 %in% unique(c(map.checks$bad.mun, map.checks$target.mun, map.checks$replace.mun))]
#
# #Applying the function to all counties (takes time)
# require(parallel)
# require(doParallel)
# no_cores <- detectCores() - 2
# cl <- makeCluster(no_cores)
# clusterExport(cl, list("possible.counties","clusterSF","country.pol"),envir=environment())
# clusterEvalQ(cl, library(sf))
# clusterEvalQ(cl, library(units))
# tmp2 <- parLapply(cl = cl, X = 1:length(possible.counties), fun = function(x) {
# state.i <- strsplit(possible.counties[x], "_")[[1]][1]
# county.i <- strsplit(possible.counties[x], "_")[[1]][2]
# check_these <- country.pol$NAME_1 %in% state.i &
# country.pol$NAME_2 %in% county.i
# polys <- country.pol[check_these,]
# clust.polys <- clusterSF(polys, 100)
# clust.polys$possible.counties[x]
# return(clust.polys)
# })
# stopImplicitCluster()
# stopCluster(cl)
# prob.counties <- possible.counties[sapply(tmp2, function(x) max(x$group) >1)]
# prob.polys <- country.pol[toto %in% prob.counties,]
# prob.data <- prob.polys
# sf::st_geometry(prob.data) <- NULL
# prob.data$loc <- toto[toto %in% prob.counties]
#
# for (i in 1:length(unique(prob.data$loc))) {
# loc <- unique(prob.data$loc)[i]
# plot_these <- prob.data$loc %in% loc
# plot(prob.polys[plot_these, 4],
# main = loc, key.pos = 1)
# # sf::st_coordinates(sf::st_centroid(prob.polys[plot_these, 4]))
# # plot(country.pol[country.pol$NAME_2 %in% prob.polys$NAME_3[plot_these],4],
# # main = loc, key.pos = 1)
# options(locatorBell=FALSE)
# locator(1)
# }
#
# resulta <- vector("list", length(possible.counties))
# names(resulta) <- possible.counties
# for (i in 1:length(possible.counties)) {
# state.i <- strsplit(possible.counties[i], "_")[[1]][1]
# county.i <- strsplit(possible.counties[i], "_")[[1]][2]
# check_these <- country.pol$NAME_1 %in% state.i &
# country.pol$NAME_2 %in% county.i
# polys <- country.pol[check_these,]
# resulta[[i]] <- clusterSF(polys, 100)
# cat(i, "\n")
# }
## Converting all maps to 'sf'
# for(i in 1:length(country.list)){
# tmp <- country.list[[i]]
# tmp1 <- sf::st_as_sf(tmp)
# country.list[[i]] <- tmp1
# }
## Inspecting
latamMap <- country.list
# latamMap_full <- country.list
#names(latamMap) <- pais
# plot(latamMap["belize"][[1]][,1])
# plot(latamMap["colombia"][[1]][,1])
# plot(latamMap["french guiana"][[1]][,1])
# plot(latamMap["paraguay"][[1]][,1])
# plot(latamMap["venezuela"][[1]][,1])
## Saving
save(latamMap, file = "./data/latamMap.rda", compress = "xz")
# save(latamMap_full, file = "./data/latamMap_full.rda", compress = "xz")
################################################################################H
################################################################################H
#####################################
### WORLD, LAND BUFFER AND SHORES ###
#####################################
#### WORLD ####
# map for function share_borders()
#world0 <- rnaturalearth::ne_download(scale = 110, type = 'countries', category = 'cultural')
world0 <- rnaturalearth::ne_download(scale = 110, type = 'map_units',
category = 'cultural')
world1 <- rgeos::gBuffer(world0, byid = TRUE, width = 0)
world1 <- cleangeo::clgeo_Clean(world1)
world2 <- rgeos::gSimplify(world1, tol = 0.001, topologyPreserve = TRUE)
world2 <- rgeos::gBuffer(world2, byid = TRUE, width = 0)
world2 <- cleangeo::clgeo_Clean(world2)
cols <- c("ISO_A2","NAME_LONG")
world3 <- sp::SpatialPolygonsDataFrame(world2, world0@data[,cols])
names(world3@data) <- c("iso_a2","name")
#Editing country name as in the gazetteer
world3@data$name <- prepCountry(world3@data$name)
#Replacing country names as in the gazetteer
tmp1 <- replaceNames[replaceNames$class %in% "country" & apply(is.na(replaceNames[, 2:4]), 1, all), ]
tmp2 <- as.character(tmp1$replace)
names(tmp2) = as.character(tmp1$pattern)
#names(tmp2) <- gsub("\\\\", "", names(tmp2))
world3@data$name <- sapply(strsplit(world3@data$name,' \\('), function(x) x[[1]][1])
world3@data$name <- stringr::str_replace_all(world3@data$name, tmp2)
#missing countries
other0 <- wo[!worldMap$NAME_0 %in% world3$name,]
names(other0@data) <- c("iso_a2","name")
other1 <- rgeos::gSimplify(other0, tol = 0.1, topologyPreserve = TRUE)
other1 <- rgeos::gBuffer(other1, byid = TRUE, width = 0)
other1 <- cleangeo::clgeo_Clean(other1)
other1 <- sp::SpatialPolygonsDataFrame(other1, other0@data)
other2 <- disaggregate(other1)
other2$area <- rgeos::gArea(other2, byid = TRUE)
other2$max.area <- NA
for (i in 1:length(unique(other2@data$name))) {
pais.i <- unique(other2$name)[i]
other2$max.area[other2$name %in% pais.i] <-
other2$area[other2$name %in% pais.i] == max(other2$area[other2$name %in% pais.i])
}
other3 <- other2[other2@data$max.area, c("iso_a2","name")]
other3 <- other3[!other3$name %in% "caspian sea",]
world4 <- rbind(world3, other3)
#Checking if all names in worldMap are in world
world4@data$name[!world4@data$name %in% gazetteer$loc[gazetteer$resolution.gazetteer %in% "country"]]
# tmp <- world3[!world3@data$name %in% gazetteer$loc[gazetteer$resolution.gazetteer %in% "country"],]
# tmp1 <- raster::coordinates(tmp)
# colnames(tmp1) <- c("x", "y")
# tmp1 <- as.data.frame(tmp1)
# sp::coordinates(tmp1) <- ~x+y
# raster::crs(tmp1) <- raster::crs(world3)
# tmp2 <- sp::SpatialPointsDataFrame(tmp1, tmp@data)
# tmp2.sf <- sf::st_as_sf(tmp2)
# tmp3 <- sf::st_intersection(tmp2.sf, worldMap)
# sf::st_geometry(tmp3) <- NULL
#Converting to sf and projecting to WSG84
world <- sf::st_as_sf(world4)
prj <- sf::st_crs(4326)
world <- sf::st_set_crs(world, prj)
#Comparing with the original wolrd map
object.size(world3)
object.size(world4)
#Repairing possible issues related to spherical geometries
isv <- sf::st_is_valid(world) # Sudan and South Georgia South Sandwich islands
if (any(isv)) {
for (i in which(!isv)) {
tmp.i <- world[i, ]$geometry
tmp.i.1 <- s2::as_s2_geography(tmp.i, check = FALSE)
tmp.i.2 <- s2::s2_union(tmp.i.1)
tmp.i.3 <- sf::st_as_sfc(
s2::s2_rebuild(tmp.i.2, s2::s2_options(split_crossing_edges = TRUE)))
world[i, ]$geometry <- tmp.i.3
}
}
# world <- sf::st_make_valid(world)
## Saving
save(world, file = "./data/world.rda", compress = "xz")
#Inspecting the maps and object sizes
# format(object.size(world0), units = "Mb") ## 15.8 Mb
# format(object.size(world2), units = "Mb") ## 15.6 Mb (not much but borders remains almost the same)
# format(object.size(world), units = "Mb") ## 10.3 Mb before; now 0.4 Mb
# sp::plot(world0, ylim=c(-24,-22), xlim=c(-45, -42))
# sp::plot(world4, add = TRUE, border = 3)
# sp::plot(world[,1], add = TRUE, border = 4)
#### LAND BUFFER ####
land <- rnaturalearth::ne_download(scale = 10, type = 'land', category = 'physical')
## Buffering mainland to 0.5 degree (~50 km at the equator) and
#buffering major islands to 0.25 degree (~25 km at the equator)
# buff.rad <- 0.5
buff.rad <- land$scalerank
buff.rad[buff.rad <= 3] <- 0.5
buff.rad[buff.rad > 3 & buff.rad < 10] <- 0.25
buff.rad[buff.rad > 10] <- 0
land.buff <- rgeos::gBuffer(land, byid=TRUE, width = buff.rad)
land.buff1 <- raster::crop(land.buff, raster::extent(CoordinateCleaner::buffland))
land.buff1 <- sp::disaggregate(land.buff1)
land.buff1 <- cleangeo::clgeo_Clean(land.buff1)
land.buff2 <- raster::aggregate(land.buff1)
land.buff3 <- rgeos::gSimplify(land.buff2, tol = 0.01, topologyPreserve = TRUE)
land.buff3 <- rgeos::gBuffer(land.buff3, byid = TRUE, width = 0)
land.buff3 <- sp::SpatialPolygonsDataFrame(land.buff3, data.frame(class = "land"))
landBuff <- sf::st_as_sf(land.buff3)
prj <- sf::st_crs(4326)
landBuff <- sf::st_set_crs(landBuff, prj)
## Saving
save(landBuff, file = "./data/landBuff.rda", compress = "xz")
#Inspecting the maps and object sizes
# sp::plot(land, ylim=c(-24,-22), xlim=c(-47, -43))
# sp::plot(land, ylim=c(15,30), xlim=c(-80, -60))
# sp::plot(land.buff2, add = TRUE, border = 2)
# sp::plot(land.buff3, add = TRUE, border = 3)
# sp::plot(landBuff, add = TRUE)
# format(object.size(land), units = "Mb") ## 12.2 Mb
# format(object.size(land.buff2), units = "Mb") ## 2.3 Mb (not much but borders remains almost the same)
# format(object.size(land.buff3), units = "Mb") ## 1.6 Mb (not much but borders remains almost the same)
# format(object.size(landBuff), units = "Mb") ## 0.9 Mb
#### MINOR ISLANDS ####
minor.islands <- rnaturalearth::ne_download(scale = 10, type = 'minor_islands', category = 'physical')
#removing islands already in the land buffer
coords <- as.data.frame(sp::coordinates(minor.islands))
colnames(coords) <- c("lon", "lat")
sp::coordinates(coords) <- ~lon + lat
sp::proj4string(coords) <- raster::crs(land)
tmp <- sp::over(coords, land.buff3)
minor.islands1 <- minor.islands[is.na(tmp$class),]
#buffering to 0.25 degree (~25 km at the equator)
buff.rad = 0.25
minor.buff <- rgeos::gBuffer(minor.islands1, byid=TRUE, width = buff.rad)
minor.buff1 <- raster::crop(minor.buff, raster::extent(CoordinateCleaner::buffland))
minor.buff1 <- sp::disaggregate(minor.buff1)
minor.buff1 <- cleangeo::clgeo_Clean(minor.buff1)
minor.buff2 <- raster::aggregate(minor.buff1)
minor.buff3 <- rgeos::gSimplify(minor.buff2, tol = 0.01, topologyPreserve = TRUE)
minor.buff3 <- rgeos::gBuffer(minor.buff3, byid = TRUE, width = 0)
minor.buff3 <- sp::SpatialPolygonsDataFrame(minor.buff3, data.frame(class = "minor.islands"))
islandsBuff <- sf::st_as_sf(minor.buff3)
prj <- sf::st_crs(4326)
islandsBuff <- sf::st_set_crs(islandsBuff, prj)
## Saving
save(islandsBuff, file = "./data/islandsBuff.rda", compress = "xz")
#Inspecting the maps and object sizes
# sp::plot(landBuff, ylim=c(-24,-22), xlim=c(-47, -43))
# sp::plot(landBuff, ylim=c(10,30), xlim=c(-90, -60))
# sp::plot(minor.buff2, add = TRUE, border = 2)
# sp::plot(minor.buff3, add = TRUE, border = 3)
# format(object.size(minor.islands), units = "Mb") ## 9.1 Mb
# format(object.size(minor.buff2), units = "Mb") ## 0.3 Mb (not much but borders remains almost the same)
# format(object.size(minor.buff3), units = "Mb") ## 0.2 Mb (not much but borders remains almost the same)
# format(object.size(islandsBuff), units = "Mb") ## 0.1 Mb
#### SHORE LINES ####
land50 <- rnaturalearth::ne_download(scale = 50, type = 'land', category = 'physical')
coast <- rnaturalearth::ne_download(scale = 50, type = 'coastline', category = 'physical')
# shore lines
land1 <- rgeos::gBuffer(land50, byid=TRUE, width=0)
land1 <- cleangeo::clgeo_Clean(land1)
land2 <- rgeos::gSimplify(land1, tol = 0.001, topologyPreserve = TRUE)
land2 <- rgeos::gBuffer(land2, byid = TRUE, width = 0)
land2 <- raster::aggregate(land2)
land3 <- as(land2, "SpatialLines")
shoreLines <- sf::st_as_sf(land3)
prj <- sf::st_crs(4326)
shoreLines <- sf::st_set_crs(shoreLines, prj)
# coast1 <- coast
# coast1@data <- coast1@data[,1,drop=FALSE]
# coast2 <- rgeos::gSimplify(coast1, tol = 0.001, topologyPreserve = TRUE)
# coastLines <- sf::st_as_sf(coast2)
# prj <- sf::st_crs(4326)
# coastLines <- sf::st_set_crs(coastLines, prj)
## Saving
save(shoreLines, file = "./data/shoreLines.rda", compress = "xz")
# save(coastLines, file = "./data/coastLines.rda", compress = "xz")
#Inspecting the maps and object sizes
# sp::plot(shoreLines, ylim=c(-25,-20), xlim=c(-47, -42))
# sp::plot(coastLines, add = TRUE, col = 2)
# format(object.size(land50), units = "Mb") ## 5.3 Mb
# format(object.size(coast), units = "Mb") ## 3.3 Mb
# format(object.size(land2), units = "Mb") ## 3.5 Mb
# format(object.size(land3), units = "Mb") ## 2.5 Mb
# format(object.size(shoreLines), units = "Mb") ##1.7 Mb
# format(object.size(coastLines), units = "Mb") ##2.3 Mb
#### A DICTIONARY OF COUNTRY NAMES AND CODES ####
# cols <- c("SOVEREIGNT","TYPE","ADMIN", "NAME","NAME_LONG","FORMAL_EN","FORMAL_FR",
# "NAME_CIAWF", "NAME_SORT","ISO_A2","ISO_A3","WB_A2","WB_A3",
# "CONTINENT", "REGION_UN", "SUBREGION", "REGION_WB",
# "NAME_DE", "NAME_EN", "NAME_ES","NAME_FR","NAME_HU","NAME_ID","NAME_IT",
# "NAME_NL","NAME_PL","NAME_PT","NAME_SV","NAME_TR","NAME_VI")
LimaRAF/plantR documentation built on Jan. 1, 2023, 10:18 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
#### Script to read shapefiles, edit and generate the data/worldMap.rds ####
require(rgdal)
require(rgeos)
require(sf)
require(sp)
require(countrycode)
devtools::load_all()
##List of Neotropical countries
countries.latam <- c("Anguilla","Antigua and Barbuda","Argentina","Aruba","Bahamas",
"Barbados","Belize","Bermuda","Bolivia","Brazil","British Virgin Islands",
"Caribbean Netherlands","Cayman Islands","Chile",
"Colombia","Costa Rica","Cuba","Curaçao","Dominica","Dominican Republic",
"Ecuador","El Salvador","Falkland Islands",
"French Guiana","Grenada","Guadeloupe","Guatemala","Guyana","Haiti",
"Honduras","Jamaica","Martinique","Mexico","Montserrat","Nicaragua",
"Panama","Paraguay","Peru","Puerto Rico","Saint Barthélemy",
"Saint Kitts and Nevis","Saint Lucia","Saint Martin (French part)","Saint Vincent and the Grenadines",
"Sint Maarten","Suriname","Trinidad and Tobago","Turks and Caicos Islands",
"United States Virgin Islands","Uruguay","Venezuela")
iso3 <- countrycode::countrycode(countries.latam, "country.name", "iso3c")
#BES = "Caribbean Netherlands" = "Bonaire, Saint Eustatius and Saba"
################################################################################H
################################################################################H
################################
### WORLD MAP FOR VALIDATION ###
################################
##new sf files
destfolder <- "vrac/latam"
destfolder <- "data-raw/GDAM"
if (!exists(destfolder)) dir.create(destfolder, recursive = T)
source("https://raw.githubusercontent.com/liibre/Rocc/master/R/getGADM.R")
#Dando problema
# purrr::walk2(.x = rep(iso3, each = 4),
# .y = rep(3:0, length(countries.latam)),
# ~ getGADM(
# cod = .x,
# level = .y,
# best = FALSE, #importante para ter outros niveis
# destfolder = destfolder
# ))
gadm_files <- list.files(destfolder, pattern = "rds$",
full.names = TRUE)
##Loading, editing and converting the world country shapefile
path <- "E://ownCloud//W_GIS"
#wo <- readOGR(dsn=paste(path,"//WO_ADM_limits_GADM36//gadm36_levels_shp",sep=""),layer="gadm36_0")
wo <- readRDS(paste(path,"//WO_ADM_limits_GADM36//gadm36_levels_shp//gadm36_0.rds",sep = ""))
#wo <- readRDS(paste0(path,"//WO_ADM_limits_GADM36//gadm36_levels_shp//gadm36_0.rds"))
#Editing country name as in the gazetteer
wo@data$pais <- countrycode(as.character(wo@data$GID_0),'iso3c', 'country.name')
wo@data$pais[is.na(wo@data$pais)] <- as.character(wo@data$NAME_0[is.na(wo@data$pais)])
wo@data$pais <- prepCountry(wo@data$pais)
# wo@data$pais <- tolower(chartr(paste(names(unwanted_array), collapse=''), paste(unwanted_array, collapse=''), wo@data$pais))
# wo@data$pais <- tolower(textclean::replace_non_ascii(wo@data$pais))
# wo@data$pais <- stringr::str_replace_all(wo@data$pais, " & ", " and ")
# wo@data$pais <- gsub("^st. ", "saint ", wo@data$pais)
# wo@data$pais <- gsub(" of the ", " ", wo@data$pais)
# wo@data$pais <- gsub(" of ", " ", wo@data$pais)
tmp1 <- replaceNames[replaceNames$class %in% "country" & apply(is.na(replaceNames[, 2:4]), 1, all), ]
tmp2 <- as.character(tmp1$replace)
names(tmp2) = as.character(tmp1$pattern)
#names(tmp2) <- gsub("\\\\", "", names(tmp2))
wo@data$pais <- sapply(strsplit(wo@data$pais,' \\('), function(x) x[[1]][1])
wo@data$pais <- stringr::str_replace_all(wo@data$pais, tmp2)
# wo@data$pais <- plantR::prepLoc(wo@data$pais)
## Comparing the map with the gazetteer
tmp <- merge(wo@data, gazetteer[,c("loc", "loc.correct")], by.x = "pais", by.y = "loc", all.x = TRUE)
tmp[!tmp$pais %in% tmp$loc.correct,] #ok!
#Transforming and simplifying
wo@data <- wo@data[,c("GID_0","pais")]
names(wo@data)[2] <- "NAME_0"
wo.simp <- gSimplify(wo, tol=0.001, topologyPreserve = TRUE)
wo.simp <- gBuffer(wo.simp, byid = TRUE, width = 0)
wo.simp1 <- SpatialPolygonsDataFrame(wo.simp, wo@data)
wo.simp1_sf <- sf::st_as_sf(wo.simp1)
# set spatial coordinates
prj <- sf::st_crs(4326)
worldMap <- sf::st_set_crs(wo.simp1_sf, prj)
# wo.simp2 <- gSimplify(wo, tol = 0.0001, topologyPreserve = TRUE)
# wo.simp2 <- gBuffer(wo.simp2, byid = TRUE, width = 0)
# wo.simp3 <- SpatialPolygonsDataFrame(wo.simp2, wo@data)
# worldMap_0001 <- sf::st_as_sf(wo.simp3)
# wo.simp4 <- gSimplify(wo, tol = 0.1, topologyPreserve = TRUE)
# wo.simp4 <- gBuffer(wo.simp4, byid = TRUE, width = 0)
# wo.simp5 <- SpatialPolygonsDataFrame(wo.simp4, wo@data)
# worldMap_01 <- sf::st_as_sf(wo.simp5)
# #wo.simp2 <- sf::st_buffer(wo.simp1, dist = 0)
# worldMap_full <- sf::st_as_sf(wo)
#
# #Inspecting
# par(mfrow = c(2,2), mar=c(0,0,1,0))
# plot(wo[1,], main="Aruba original")
# plot(wo.simp[1,], border = "red", main="Aruba gSimp 0.001")
# plot(st_geometry(worldMap[1,1]), border = "green", main = "Aruba gSimp 0.001 sf")
#Repairing possible issues related to spherical geometries
isv <- sf::st_is_valid(worldMap) #Senegal had an issue
if (any(isv)) {
for (i in which(!isv)) {
tmp.i <- worldMap[i, ]$geometry
tmp.i.1 <- s2::as_s2_geography(sf::st_as_binary(tmp.i), check = FALSE)
tmp.i.2 <- s2::s2_as_binary(s2::s2_snap_to_grid(tmp.i.1, grid_size = 1e-7))
tmp.i.3 <- sf::st_as_sfc(s2::s2_geog_from_wkb(tmp.i.2, check = TRUE))
worldMap[i, ]$geometry <- tmp.i.3
}
}
# worldMap <- sf::st_make_valid(worldMap)
#Saving
save(worldMap, file = "./data/worldMap.rda", compress = "xz")
# save(worldMap_0001, file = "./data/worldMap_0001.rda", compress = "xz")
# save(worldMap_full, file = "./data/worldMap_full.rda", compress = "xz")
################################################################################H
################################################################################H
#########################################
### LATIN AMERICAN MAP FOR VALIDATION ###
#########################################
##Loading, editing and converting the world country shapefile
path0 <- "E://ownCloud//W_GIS//Am_Lat_ADM_GADM_v3.6//"
countries <- list.files(path0, full.name=TRUE)
countries <- countries[grep(paste(iso3, collapse = "|") ,countries)]
countries <- countries[grep("sf.rds",countries)]
countries3 <- countries[grep('_3_sf.rds',countries)]
adm3 <- sapply(strsplit(countries3, "//|_"), function(x) x[11])
countries2 <- countries[grepl('_2_sf.rds',countries) & !grepl(paste0(c(adm3), collapse = "|"), countries)]
adm2 <- sapply(strsplit(countries2, "//|_"), function(x) x[11])
countries1 <- countries[grepl('_1_sf.rds',countries) & !grepl(paste0(c(adm2,adm3), collapse = "|"), countries)]
adm1 <- sapply(strsplit(countries1, "//|_"), function(x) x[11])
countries0 <- countries[grepl('_0_sf.rds',countries) & !grepl(paste0(c(adm1,adm2,adm3), collapse = "|"), countries)]
adm0 <- sapply(strsplit(countries0, "//|_"), function(x) x[11])
all.countries <- sort(c(adm1, adm2, adm3))
iso3[!iso3 %in% all.countries] # all of these countries are only available at ADM_0, that is, already in worldMap.
countries <- sort(c(countries1, countries2, countries3))
country.list <- vector('list', length(countries))
pais = NULL
# wo <- readRDS(gadm_files)
# wo <- purrr::map(gadm_files, .f = function(x) readRDS(file = x))
for (i in 1:length(country.list)) {
path = countries[i]
tmp = readRDS(path)
#tmp1 = tmp@data
tmp1 = tmp[,grepl("^NAME_", names(tmp))]
#country
tmp1$NAME_0 <- prepCountry(tmp1$NAME_0)
# tmp1$NAME_0 <- tolower(textclean::replace_non_ascii(tmp1$NAME_0))
# tmp1$NAME_0 <- stringr::str_replace_all(tmp1$NAME_0, " & ", " and ")
# tmp1$NAME_0 <- gsub("^st. ", "saint ", tmp1$NAME_0)
# tmp1$NAME_0 <- gsub(" of the ", " ", tmp1$NAME_0)
# tmp1$NAME_0 <- gsub(" of ", " ", tmp1$NAME_0)
toto1 <- replaceNames[replaceNames$class %in% "country" & apply(is.na(replaceNames[, 2:4]), 1, all), ]
toto2 <- as.character(toto1$replace)
names(toto2) = as.character(toto1$pattern)
# names(toto2) <- gsub("\\\\", "", names(toto2))
tmp1$NAME_0 <- sapply(strsplit(tmp1$NAME_0,' \\('), function(x) x[[1]][1])
tmp1$NAME_0 <- stringr::str_replace_all(tmp1$NAME_0, toto2)
# tmp1$NAME_0 <- plantR::prepLoc(tmp1$NAME_0)
tmp1$NAME_0 <- gsub("bonaire, sint eustatius and saba", "caribbean netherlands", tmp1$NAME_0)
tmp1$NAME_0 <- gsub("bonaire, sint eustatius saba", "caribbean netherlands", tmp1$NAME_0)
tmp1$NAME_0 <- gsub("saint vincent and the grenadines", "saint vincent grenadines", tmp1$NAME_0)
tmp1$NAME_0 <- gsub("virgin islands, u.s.", "united states virgin islands", tmp1$NAME_0)
#ADM_1
tmp1$NAME_1 <- tolower(textclean::replace_non_ascii(tmp1$NAME_1))
tmp1$NAME_1 <- plantR::prepLoc(tmp1$NAME_1)
if ("NAME_2" %in% names(tmp1)) {
tmp1$NAME_2 <- tolower(textclean::replace_non_ascii(tmp1$NAME_2))
tmp1$NAME_2 <- plantR::prepLoc(tmp1$NAME_2)
}
if ("NAME_3" %in% names(tmp1)) {
tmp1$NAME_3 <- tolower(textclean::replace_non_ascii(tmp1$NAME_3))
tmp1$NAME_3 <- plantR::prepLoc(tmp1$NAME_3)
}
#Symplifying the maps
# tmp.simp <- gSimplify(tmp, tol=0.0001, topologyPreserve = TRUE)
# tmp.simp <- gBuffer(tmp.simp, byid=TRUE, width=0)
tmp.simp <- st_simplify(tmp1, dTolerance = 0.0001, preserveTopology = TRUE)
tmp.simp <- st_buffer(tmp.simp, dist=0)
# tmp.simp <- tmp1
##Simplify transform MULTIPOLYGONS in POLYGON: is that a problem?
#Extract polygon ID's
# pid <- sapply(slot(tmp.simp, "polygons"), function(x) slot(x, "ID"))
#
# #Create dataframe with correct rownames and merging
# p.df <- as.data.frame(tmp1, row.names = pid, stringsAsFactors = FALSE)
# tmp.simp1 <- SpatialPolygonsDataFrame(tmp.simp, p.df)
# set spatial coordinates
prj <- sf::st_crs(4326)
tmp.simp <- sf::st_set_crs(tmp.simp, prj)
#Saving
# pais[i] = unique(p.df$NAME_0)
pais[i] = unique(tmp.simp$NAME_0)
country.list[[i]] = tmp.simp
}
names(country.list) = pais
#pais[!pais %in% wo.simp1$NAME_0]
## Standardizing the shapefile and gazetteer names (for Brazil only)
# #loading the gazetteer and removing possible duplicated localities
# dic <- read.csv("data-raw//raw_dictionaries//gazetteer.csv",as.is=TRUE)
# # dic <- read.csv("C://Users//renato//Documents//raflima//Pos Doc//Manuscritos//Artigo AF checklist//data analysis//dictionaries//gazetteer.csv",as.is=TRUE)
# dic <- dic[dic$status %in% "ok",]
# dic <- dic[dic$resolution.gazetteer %in% c("country","state","county"),]
#
# j = which(names(country.list) == "brazil")
# #for(j in 31:length(country.list)) {
# #Creating the locality strings for the shapefile dataframe
# tmp = country.list[[j]]#@data
# #for(i in 1:dim(tmp)[2]) tmp[,i] <- as.character(tmp[,i])
#
# tmp.df <- tmp
# st_geometry(tmp.df) <- NULL
# # tmp1= data.frame(loc=apply(tmp, 1, paste, collapse="_"))
# ids <- names(tmp.df) %in% c("NAME_0","NAME_1","NAME_2")
# tmp1= data.frame(loc=apply(tmp.df[, ids], 1, paste, collapse="_"),
# stringsAsFactors = FALSE)
# # tmp1$loc = as.character(tmp1$loc)
#
# tmp$order = 1:dim(tmp)[1]
# tmp1$order = 1:dim(tmp1)[1]
#
# #merging the two sources of info
# tmp2 <- merge(tmp1, dic[,c("NAME_0","NAME_1","NAME_2","loc","loc.correct")],
# by="loc", all.x=TRUE)
# tmp3 <- strsplit(tmp2$loc.correct,"_")
#
# nms <- c("country", "state", "county")
# for(w in 1:sapply(tmp3, length)[1]) {
# n <- rep(NA, dim(tmp2)[1])
# n <- sapply(tmp3, function(x) x[w])
# df <- data.frame(n, stringsAsFactors = FALSE); names(df) = nms[w]
# tmp2 <- cbind.data.frame(tmp2, df, stringsAsFactors = FALSE)
# }
#
# #comparing and replacing the correct info
# tmp2 <- tmp2[match(tmp$order, tmp2$order),]
# table(tmp$order == tmp2$order)
# id = !tmp2$loc %in% tmp2$loc.correct
#
# if(any(id)) {
# cat(paste(i,": ",unique(tmp$NAME_0),table(id)),"\n")
#
# table(tmp$NAME_0[id] == tmp2[id,c("country")])
# #tmp[id,1] = tmp2[id,c("country")]
# table(tmp$NAME_1[id] == tmp2[id,c("state")]) #all differences were double-checked and the gazetteer is right!
# cbind(tmp$NAME_1[id], tmp2[id,c("state","county")])[tmp$NAME_1[id] == tmp2[id,c("state")],]
# #tmp[id,2] = tmp2[id,c("state")]
# table(!tmp$NAME_2[id] == tmp2[id,c("county")]) #all differences were double-checked and the gazetteer is right!
#
# cbind(tmp$NAME_1[id],tmp$NAME_2[id], tmp2$county[id])[!tmp$NAME_2[id] == tmp2$county[id],]
# # tmp$NAME_2[id] = tmp2[id,c("county")]
#
# #Saving the changes
# country.list[[j]]$NAME_2[id] <- tmp2$county[id]
# # country.list[[j]]@data = tmp[,c("NAME_0","NAME_1","NAME_2")]
# }
# #}
# ## Other GADM issues and mixed names
# #Manaus
# tmp <- country.list[[j]][country.list[[j]]$NAME_2 %in% "maues",]
# tmp1 <- sf::st_coordinates(sf::st_centroid(tmp))
# ids.manaus <- tmp1[,1] < (-59) & tmp1[,2] > (-3.5)
# country.list[[j]]$NAME_2[country.list[[j]]$NAME_2 %in% "maues"][ids.manaus] <- "manaus"
#
# country.list[[j]]$check[country.list[[j]]$NAME_2 %in% "maues"] <- ids.manaus
#
# #Brasilândia de Minas/Brazópolis
# replace_this <- country.list[[j]]$NAME_2 %in% "brazopolis" &
# country.list[[j]]$NAME_3 %in% "brazilandia minas"
# country.list[[j]]$NAME_2[replace_this] <- "brasilandia minas"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #LEM
# replace_this <- country.list[[j]]$NAME_2 %in% "barreiras" &
# country.list[[j]]$NAME_3 %in% "luis eduardo magalhaes"
# country.list[[j]]$NAME_2[replace_this] <- "luis eduardo magalhaes"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Barrocas
# replace_this <- country.list[[j]]$NAME_2 %in% "serrinha" &
# country.list[[j]]$NAME_3 %in% "barrocas"
# country.list[[j]]$NAME_2[replace_this] <- "barrocas"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Capao Bonito do Sul
# replace_this <- country.list[[j]]$NAME_2 %in% "lagoa vermelha" &
# country.list[[j]]$NAME_3 %in% "capao bonito"
# country.list[[j]]$NAME_2[replace_this] <- "capao bonito sul"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Santa Cecília do Sul
# replace_this <- country.list[[j]]$NAME_2 %in% "tapejara" &
# country.list[[j]]$NAME_3 %in% "santa cecilia"
# country.list[[j]]$NAME_2[replace_this] <- "santa cecilia sul"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Balneario Rincão/SC
# replace_this <- country.list[[j]]$NAME_2 %in% "icara" &
# country.list[[j]]$NAME_3 %in% "balnerio rincao"
# country.list[[j]]$NAME_2[replace_this] <- "balneario rincao"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Três Barras/SC
# replace_this <- country.list[[j]]$NAME_2 %in% "sao mateus sul" &
# country.list[[j]]$NAME_3 %in% "tres barras"
# country.list[[j]]$NAME_1[replace_this] <- "santa catarina"
# country.list[[j]]$NAME_2[replace_this] <- "tres barras"
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
#
# #Teixeira Soares
# replace_this <- country.list[[j]]$NAME_2 %in% "texeira soares"
# country.list[[j]]$NAME_2[replace_this] <- "teixeira soares"
#
# #Mato Grosso as a county
# replace_this <- country.list[[j]]$NAME_2 %in% "mato grosso" &
# country.list[[j]]$NAME_3 %in% "aguapei"
# country.list[[j]]$NAME_2[replace_this] <- "vila bela santissima trindade"
#
# #Itabirinha and Itabirito/MG
# replace_this <- country.list[[j]]$NAME_2 %in% "itabirito" &
# country.list[[j]]$NAME_3 %in% "itabirinha mantena"
# country.list[[j]]$NAME_2[replace_this] <- "itabirinha"
#
#
# #Other counties
# map.checks <- readRDS("./data-raw/map.replacements.rds")
# paises <- sapply(map.checks$locs, function (x) strsplit(x, "_")[[1]][1])
# estados <- sapply(map.checks$locs, function (x) strsplit(x, "_")[[1]][2])
# j = which(names(country.list) == "brazil")
# for(w in 1:length(map.checks$locs)) {
# pais.i <- paises[w]
# estado.i <- estados[w]
# bad <- map.checks$municipio.bad[w]
# good <- map.checks$municipio.good[w]
# replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
# country.list[[j]]$NAME_1 %in% estado.i &
# country.list[[j]]$NAME_2 %in% bad &
# country.list[[j]]$NAME_3 %in% good
# country.list[[j]]$NAME_2[replace_this] <- good
# country.list[[j]]$NAME_3[replace_this] <- NA_character_
# }
# Fixing GADM problems
map.checks <- read.csv("./data-raw/raw_dictionaries/gadmCheck.csv")
j = which(names(country.list) == "brazil")
for(w in 1:dim(map.checks)[1]) {
pais.i <- map.checks$pais[w]
estado.i <- map.checks$estado[w]
bad.i <- map.checks$bad.mun[w]
target.i <- map.checks$target.mun[w]
replace.i <- map.checks$replace.mun[w]
case.i <- map.checks$case[w]
if (case.i == "first") {
replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
country.list[[j]]$NAME_1 %in% estado.i &
country.list[[j]]$NAME_2 %in% bad.i &
country.list[[j]]$NAME_3 %in% target.i
if (any(replace_this))
country.list[[j]]$NAME_2[replace_this] <- replace.i
}
if (case.i == "second") {
replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
country.list[[j]]$NAME_1 %in% estado.i &
country.list[[j]]$NAME_2 %in% bad.i
if (any(replace_this))
country.list[[j]]$NAME_2[replace_this] <- replace.i
}
if (case.i == "county_state") {
replace.i.1 <- strsplit(replace.i,"\\|")[[1]][1]
replace.i.2 <- strsplit(replace.i,"\\|")[[1]][2]
replace_this <- country.list[[j]]$NAME_0 %in% pais.i &
country.list[[j]]$NAME_1 %in% estado.i &
country.list[[j]]$NAME_2 %in% bad.i &
country.list[[j]]$NAME_3 %in% target.i
if (any(replace_this)) {
country.list[[j]]$NAME_1[replace_this] <- replace.i.1
country.list[[j]]$NAME_2[replace_this] <- replace.i.2
}
}
}
# ## Finding possible problems in GADM polygons: counties with disjoint polygons
# #Function that aggreagte polygons sharing their borders
# clusterSF <- function(sfpolys, thresh){
# df <- sfpolys[, grepl("NAME_", names(sfpolys))]
# sf::st_geometry(df) <- NULL
# nome <- unique(apply(df[,1:3], 1, paste, collapse = "_"))[1]
#
# dmat <- sf::st_distance(sfpolys)
# thresh.dist <- units::as_units('m', thresh)
# hc <- stats::hclust(as.dist(dmat > thresh.dist), method="single")
#
# groups = stats::cutree(hc, h=0.5)
# d = sf::st_sf(
# geom = do.call(c,
# lapply(1:max(groups), function(g){
# sf::st_union(sfpolys[groups==g,])
# })
# )
# )
# d$name <- nome
# d$group <- 1:nrow(d)
# d
# }
#
# #Selected country map
# country.pol <- country.list[[j]]
# # country.pol <- country.list[[which(names(country.list) == "brazil")]]
#
# #List of counties with more than one polygon
# toto <- paste(country.pol$NAME_1,country.pol$NAME_2, sep="_")
# possible.counties <- unique(toto[duplicated(toto)])
# #possible.counties <- possible.counties[country.pol$NAME_2 %in% unique(c(map.checks$bad.mun, map.checks$target.mun, map.checks$replace.mun))]
#
# #Applying the function to all counties (takes time)
# require(parallel)
# require(doParallel)
# no_cores <- detectCores() - 2
# cl <- makeCluster(no_cores)
# clusterExport(cl, list("possible.counties","clusterSF","country.pol"),envir=environment())
# clusterEvalQ(cl, library(sf))
# clusterEvalQ(cl, library(units))
# tmp2 <- parLapply(cl = cl, X = 1:length(possible.counties), fun = function(x) {
# state.i <- strsplit(possible.counties[x], "_")[[1]][1]
# county.i <- strsplit(possible.counties[x], "_")[[1]][2]
# check_these <- country.pol$NAME_1 %in% state.i &
# country.pol$NAME_2 %in% county.i
# polys <- country.pol[check_these,]
# clust.polys <- clusterSF(polys, 100)
# clust.polys$possible.counties[x]
# return(clust.polys)
# })
# stopImplicitCluster()
# stopCluster(cl)
# prob.counties <- possible.counties[sapply(tmp2, function(x) max(x$group) >1)]
# prob.polys <- country.pol[toto %in% prob.counties,]
# prob.data <- prob.polys
# sf::st_geometry(prob.data) <- NULL
# prob.data$loc <- toto[toto %in% prob.counties]
#
# for (i in 1:length(unique(prob.data$loc))) {
# loc <- unique(prob.data$loc)[i]
# plot_these <- prob.data$loc %in% loc
# plot(prob.polys[plot_these, 4],
# main = loc, key.pos = 1)
# # sf::st_coordinates(sf::st_centroid(prob.polys[plot_these, 4]))
# # plot(country.pol[country.pol$NAME_2 %in% prob.polys$NAME_3[plot_these],4],
# # main = loc, key.pos = 1)
# options(locatorBell=FALSE)
# locator(1)
# }
#
# resulta <- vector("list", length(possible.counties))
# names(resulta) <- possible.counties
# for (i in 1:length(possible.counties)) {
# state.i <- strsplit(possible.counties[i], "_")[[1]][1]
# county.i <- strsplit(possible.counties[i], "_")[[1]][2]
# check_these <- country.pol$NAME_1 %in% state.i &
# country.pol$NAME_2 %in% county.i
# polys <- country.pol[check_these,]
# resulta[[i]] <- clusterSF(polys, 100)
# cat(i, "\n")
# }
## Converting all maps to 'sf'
# for(i in 1:length(country.list)){
# tmp <- country.list[[i]]
# tmp1 <- sf::st_as_sf(tmp)
# country.list[[i]] <- tmp1
# }
## Inspecting
latamMap <- country.list
# latamMap_full <- country.list
#names(latamMap) <- pais
# plot(latamMap["belize"][[1]][,1])
# plot(latamMap["colombia"][[1]][,1])
# plot(latamMap["french guiana"][[1]][,1])
# plot(latamMap["paraguay"][[1]][,1])
# plot(latamMap["venezuela"][[1]][,1])
## Saving
save(latamMap, file = "./data/latamMap.rda", compress = "xz")
# save(latamMap_full, file = "./data/latamMap_full.rda", compress = "xz")
################################################################################H
################################################################################H
#####################################
### WORLD, LAND BUFFER AND SHORES ###
#####################################
#### WORLD ####
# map for function share_borders()
#world0 <- rnaturalearth::ne_download(scale = 110, type = 'countries', category = 'cultural')
world0 <- rnaturalearth::ne_download(scale = 110, type = 'map_units',
category = 'cultural')
world1 <- rgeos::gBuffer(world0, byid = TRUE, width = 0)
world1 <- cleangeo::clgeo_Clean(world1)
world2 <- rgeos::gSimplify(world1, tol = 0.001, topologyPreserve = TRUE)
world2 <- rgeos::gBuffer(world2, byid = TRUE, width = 0)
world2 <- cleangeo::clgeo_Clean(world2)
cols <- c("ISO_A2","NAME_LONG")
world3 <- sp::SpatialPolygonsDataFrame(world2, world0@data[,cols])
names(world3@data) <- c("iso_a2","name")
#Editing country name as in the gazetteer
world3@data$name <- prepCountry(world3@data$name)
#Replacing country names as in the gazetteer
tmp1 <- replaceNames[replaceNames$class %in% "country" & apply(is.na(replaceNames[, 2:4]), 1, all), ]
tmp2 <- as.character(tmp1$replace)
names(tmp2) = as.character(tmp1$pattern)
#names(tmp2) <- gsub("\\\\", "", names(tmp2))
world3@data$name <- sapply(strsplit(world3@data$name,' \\('), function(x) x[[1]][1])
world3@data$name <- stringr::str_replace_all(world3@data$name, tmp2)
#missing countries
other0 <- wo[!worldMap$NAME_0 %in% world3$name,]
names(other0@data) <- c("iso_a2","name")
other1 <- rgeos::gSimplify(other0, tol = 0.1, topologyPreserve = TRUE)
other1 <- rgeos::gBuffer(other1, byid = TRUE, width = 0)
other1 <- cleangeo::clgeo_Clean(other1)
other1 <- sp::SpatialPolygonsDataFrame(other1, other0@data)
other2 <- disaggregate(other1)
other2$area <- rgeos::gArea(other2, byid = TRUE)
other2$max.area <- NA
for (i in 1:length(unique(other2@data$name))) {
pais.i <- unique(other2$name)[i]
other2$max.area[other2$name %in% pais.i] <-
other2$area[other2$name %in% pais.i] == max(other2$area[other2$name %in% pais.i])
}
other3 <- other2[other2@data$max.area, c("iso_a2","name")]
other3 <- other3[!other3$name %in% "caspian sea",]
world4 <- rbind(world3, other3)
#Checking if all names in worldMap are in world
world4@data$name[!world4@data$name %in% gazetteer$loc[gazetteer$resolution.gazetteer %in% "country"]]
# tmp <- world3[!world3@data$name %in% gazetteer$loc[gazetteer$resolution.gazetteer %in% "country"],]
# tmp1 <- raster::coordinates(tmp)
# colnames(tmp1) <- c("x", "y")
# tmp1 <- as.data.frame(tmp1)
# sp::coordinates(tmp1) <- ~x+y
# raster::crs(tmp1) <- raster::crs(world3)
# tmp2 <- sp::SpatialPointsDataFrame(tmp1, tmp@data)
# tmp2.sf <- sf::st_as_sf(tmp2)
# tmp3 <- sf::st_intersection(tmp2.sf, worldMap)
# sf::st_geometry(tmp3) <- NULL
#Converting to sf and projecting to WSG84
world <- sf::st_as_sf(world4)
prj <- sf::st_crs(4326)
world <- sf::st_set_crs(world, prj)
#Comparing with the original wolrd map
object.size(world3)
object.size(world4)
#Repairing possible issues related to spherical geometries
isv <- sf::st_is_valid(world) # Sudan and South Georgia South Sandwich islands
if (any(isv)) {
for (i in which(!isv)) {
tmp.i <- world[i, ]$geometry
tmp.i.1 <- s2::as_s2_geography(tmp.i, check = FALSE)
tmp.i.2 <- s2::s2_union(tmp.i.1)
tmp.i.3 <- sf::st_as_sfc(
s2::s2_rebuild(tmp.i.2, s2::s2_options(split_crossing_edges = TRUE)))
world[i, ]$geometry <- tmp.i.3
}
}
# world <- sf::st_make_valid(world)
## Saving
save(world, file = "./data/world.rda", compress = "xz")
#Inspecting the maps and object sizes
# format(object.size(world0), units = "Mb") ## 15.8 Mb
# format(object.size(world2), units = "Mb") ## 15.6 Mb (not much but borders remains almost the same)
# format(object.size(world), units = "Mb") ## 10.3 Mb before; now 0.4 Mb
# sp::plot(world0, ylim=c(-24,-22), xlim=c(-45, -42))
# sp::plot(world4, add = TRUE, border = 3)
# sp::plot(world[,1], add = TRUE, border = 4)
#### LAND BUFFER ####
land <- rnaturalearth::ne_download(scale = 10, type = 'land', category = 'physical')
## Buffering mainland to 0.5 degree (~50 km at the equator) and
#buffering major islands to 0.25 degree (~25 km at the equator)
# buff.rad <- 0.5
buff.rad <- land$scalerank
buff.rad[buff.rad <= 3] <- 0.5
buff.rad[buff.rad > 3 & buff.rad < 10] <- 0.25
buff.rad[buff.rad > 10] <- 0
land.buff <- rgeos::gBuffer(land, byid=TRUE, width = buff.rad)
land.buff1 <- raster::crop(land.buff, raster::extent(CoordinateCleaner::buffland))
land.buff1 <- sp::disaggregate(land.buff1)
land.buff1 <- cleangeo::clgeo_Clean(land.buff1)
land.buff2 <- raster::aggregate(land.buff1)
land.buff3 <- rgeos::gSimplify(land.buff2, tol = 0.01, topologyPreserve = TRUE)
land.buff3 <- rgeos::gBuffer(land.buff3, byid = TRUE, width = 0)
land.buff3 <- sp::SpatialPolygonsDataFrame(land.buff3, data.frame(class = "land"))
landBuff <- sf::st_as_sf(land.buff3)
prj <- sf::st_crs(4326)
landBuff <- sf::st_set_crs(landBuff, prj)
## Saving
save(landBuff, file = "./data/landBuff.rda", compress = "xz")
#Inspecting the maps and object sizes
# sp::plot(land, ylim=c(-24,-22), xlim=c(-47, -43))
# sp::plot(land, ylim=c(15,30), xlim=c(-80, -60))
# sp::plot(land.buff2, add = TRUE, border = 2)
# sp::plot(land.buff3, add = TRUE, border = 3)
# sp::plot(landBuff, add = TRUE)
# format(object.size(land), units = "Mb") ## 12.2 Mb
# format(object.size(land.buff2), units = "Mb") ## 2.3 Mb (not much but borders remains almost the same)
# format(object.size(land.buff3), units = "Mb") ## 1.6 Mb (not much but borders remains almost the same)
# format(object.size(landBuff), units = "Mb") ## 0.9 Mb
#### MINOR ISLANDS ####
minor.islands <- rnaturalearth::ne_download(scale = 10, type = 'minor_islands', category = 'physical')
#removing islands already in the land buffer
coords <- as.data.frame(sp::coordinates(minor.islands))
colnames(coords) <- c("lon", "lat")
sp::coordinates(coords) <- ~lon + lat
sp::proj4string(coords) <- raster::crs(land)
tmp <- sp::over(coords, land.buff3)
minor.islands1 <- minor.islands[is.na(tmp$class),]
#buffering to 0.25 degree (~25 km at the equator)
buff.rad = 0.25
minor.buff <- rgeos::gBuffer(minor.islands1, byid=TRUE, width = buff.rad)
minor.buff1 <- raster::crop(minor.buff, raster::extent(CoordinateCleaner::buffland))
minor.buff1 <- sp::disaggregate(minor.buff1)
minor.buff1 <- cleangeo::clgeo_Clean(minor.buff1)
minor.buff2 <- raster::aggregate(minor.buff1)
minor.buff3 <- rgeos::gSimplify(minor.buff2, tol = 0.01, topologyPreserve = TRUE)
minor.buff3 <- rgeos::gBuffer(minor.buff3, byid = TRUE, width = 0)
minor.buff3 <- sp::SpatialPolygonsDataFrame(minor.buff3, data.frame(class = "minor.islands"))
islandsBuff <- sf::st_as_sf(minor.buff3)
prj <- sf::st_crs(4326)
islandsBuff <- sf::st_set_crs(islandsBuff, prj)
## Saving
save(islandsBuff, file = "./data/islandsBuff.rda", compress = "xz")
#Inspecting the maps and object sizes
# sp::plot(landBuff, ylim=c(-24,-22), xlim=c(-47, -43))
# sp::plot(landBuff, ylim=c(10,30), xlim=c(-90, -60))
# sp::plot(minor.buff2, add = TRUE, border = 2)
# sp::plot(minor.buff3, add = TRUE, border = 3)
# format(object.size(minor.islands), units = "Mb") ## 9.1 Mb
# format(object.size(minor.buff2), units = "Mb") ## 0.3 Mb (not much but borders remains almost the same)
# format(object.size(minor.buff3), units = "Mb") ## 0.2 Mb (not much but borders remains almost the same)
# format(object.size(islandsBuff), units = "Mb") ## 0.1 Mb
#### SHORE LINES ####
land50 <- rnaturalearth::ne_download(scale = 50, type = 'land', category = 'physical')
coast <- rnaturalearth::ne_download(scale = 50, type = 'coastline', category = 'physical')
# shore lines
land1 <- rgeos::gBuffer(land50, byid=TRUE, width=0)
land1 <- cleangeo::clgeo_Clean(land1)
land2 <- rgeos::gSimplify(land1, tol = 0.001, topologyPreserve = TRUE)
land2 <- rgeos::gBuffer(land2, byid = TRUE, width = 0)
land2 <- raster::aggregate(land2)
land3 <- as(land2, "SpatialLines")
shoreLines <- sf::st_as_sf(land3)
prj <- sf::st_crs(4326)
shoreLines <- sf::st_set_crs(shoreLines, prj)
# coast1 <- coast
# coast1@data <- coast1@data[,1,drop=FALSE]
# coast2 <- rgeos::gSimplify(coast1, tol = 0.001, topologyPreserve = TRUE)
# coastLines <- sf::st_as_sf(coast2)
# prj <- sf::st_crs(4326)
# coastLines <- sf::st_set_crs(coastLines, prj)
## Saving
save(shoreLines, file = "./data/shoreLines.rda", compress = "xz")
# save(coastLines, file = "./data/coastLines.rda", compress = "xz")
#Inspecting the maps and object sizes
# sp::plot(shoreLines, ylim=c(-25,-20), xlim=c(-47, -42))
# sp::plot(coastLines, add = TRUE, col = 2)
# format(object.size(land50), units = "Mb") ## 5.3 Mb
# format(object.size(coast), units = "Mb") ## 3.3 Mb
# format(object.size(land2), units = "Mb") ## 3.5 Mb
# format(object.size(land3), units = "Mb") ## 2.5 Mb
# format(object.size(shoreLines), units = "Mb") ##1.7 Mb
# format(object.size(coastLines), units = "Mb") ##2.3 Mb
#### A DICTIONARY OF COUNTRY NAMES AND CODES ####
# cols <- c("SOVEREIGNT","TYPE","ADMIN", "NAME","NAME_LONG","FORMAL_EN","FORMAL_FR",
# "NAME_CIAWF", "NAME_SORT","ISO_A2","ISO_A3","WB_A2","WB_A3",
# "CONTINENT", "REGION_UN", "SUBREGION", "REGION_WB",
# "NAME_DE", "NAME_EN", "NAME_ES","NAME_FR","NAME_HU","NAME_ID","NAME_IT",
# "NAME_NL","NAME_PL","NAME_PT","NAME_SV","NAME_TR","NAME_VI")
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.