.Rhistory

res <- cbind.data.frame(tax = EOO.poly.name[x],
prop,
stringsAsFactors = FALSE)
if(output_raster %in% "area.table")
area <- round((prop * r.area)/100, 3)
res <- cbind.data.frame(tax = EOO.poly.name[x],
area,
stringsAsFactors = FALSE)
}
}
}
res
# if(is.null(result)) result <- res else result <- rbind.data.frame(result, res, stringsAsFactors = FALSE)
# cat(x, "\n")
}
if(parallel) snow::stopCluster(cl)
if(show_progress) close(pb)
result <- do.call(rbind.data.frame, output)
rownames(result) <- NULL
return(result)
}
EOO.habitat(EOO.poly[1:10,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 4)
require(ConR)
EOO.habitat(EOO.poly[1:10,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 4)
## EOO shapefiles
EOO.poly <- readRDS(file = "C:/Users/renato/Documents/raflima/Pos Doc/Manuscritos/Artigo Extincao na MA/data analysis/data/spp.convex.hull.polys_sf_uncropped.rds")
### HUMAN INFLUENCES ### between 1995 - 2004 (0 low and 65 is high)
hab.map <- raster::raster(paste(path,"//Am_Lat_Global_Human_Influence_Index_v2//hii_global//hii_neotrop.tif",sep=""))
## ESA land cover
path <- "E://ownCloud//W_GIS" # Renato's path
### HUMAN INFLUENCES ### between 1995 - 2004 (0 low and 65 is high)
hab.map <- raster::raster(paste(path,"//Am_Lat_Global_Human_Influence_Index_v2//hii_global//hii_neotrop.tif",sep=""))
image(hab.map)
raster::image(hab.map)
EOO.habitat(EOO.poly[1:10,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 4)
toto2.1 <- EOO.habitat(EOO.poly[1:50,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 4)
x=1
proj_crs <- proj_crs(proj_type = proj_type)
proj_hab <- raster::crs(hab.map)
if(is.na(sf::st_crs(EOO.poly)))
sf::st_crs(EOO.poly) <- 4326
EOO.poly <- sf::st_transform(EOO.poly, crs = proj_hab)
# Croping EOO shapefiles to the raster extent
EOO.poly.crop <- suppressWarnings(
sf::st_intersection(EOO.poly, sf::st_as_sfc(sf::st_bbox(hab.map))))
if (length(EOO.poly$geometry) > length(EOO.poly.crop$geometry))
warning(paste0("The EOO of ",
length(EOO.poly$geometry) - length(EOO.poly.crop$geometry),
" species became empty after croping them to the habitat map extent"))
EOO.poly.proj <- sf::st_transform(EOO.poly.crop, crs = proj_crs)
EOO.poly.area <- as.double(sf::st_area(EOO.poly.proj))/1000000 # area in km2
EOO.poly.name <- EOO.poly.crop
sf::st_geometry(EOO.poly.name) <- NULL
EOO.poly.name <- EOO.poly.name[,1]
if(any(grepl("SpatialPolygon", class(hab.map))))
hab.map <- sf::st_as_sf(hab.map)
if (is.null(hab.class) & any(grepl("Raster", class(hab.map))))
classes <- seq(hab.map[[1]]@data@min, hab.map[[1]]@data@max, 1)
### HUMAN INFLUENCES ### between 1995 - 2004 (0 low and 65 is high)
hab.map <- raster::raster(paste(path,"//Am_Lat_Global_Human_Influence_Index_v2//hii_global//hii_neotrop.tif",sep=""))
hab.class = NULL
years = NULL
export_shp = FALSE
plot = FALSE
output_raster = "summary"
proj_type = "cea"
parallel = TRUE
NbeCores = 4
show_progress = TRUE
proj_crs <- proj_crs(proj_type = proj_type)
proj_hab <- raster::crs(hab.map)
if(is.na(sf::st_crs(EOO.poly)))
sf::st_crs(EOO.poly) <- 4326
EOO.poly <- sf::st_transform(EOO.poly, crs = proj_hab)
# Croping EOO shapefiles to the raster extent
EOO.poly.crop <- suppressWarnings(
sf::st_intersection(EOO.poly, sf::st_as_sfc(sf::st_bbox(hab.map))))
if (length(EOO.poly$geometry) > length(EOO.poly.crop$geometry))
warning(paste0("The EOO of ",
length(EOO.poly$geometry) - length(EOO.poly.crop$geometry),
" species became empty after croping them to the habitat map extent"))
EOO.poly.proj <- sf::st_transform(EOO.poly.crop, crs = proj_crs)
EOO.poly.area <- as.double(sf::st_area(EOO.poly.proj))/1000000 # area in km2
EOO.poly.name <- EOO.poly.crop
sf::st_geometry(EOO.poly.name) <- NULL
EOO.poly.name <- EOO.poly.name[,1]
if(any(grepl("SpatialPolygon", class(hab.map))))
hab.map <- sf::st_as_sf(hab.map)
if (is.null(hab.class) & any(grepl("Raster", class(hab.map))))
classes <- seq(hab.map[[1]]@data@min, hab.map[[1]]@data@max, 1)
x
any(grepl("sf", class(hab.map)))
grepl("Raster", class(hab.map))
ext <- raster::extent(sf::st_bbox(EOO.poly.crop[x, ])[c(1,3,2,4)])
crop <- raster::crop(hab.map, ext) # cropping raster to the extent of the polygon
mask <- raster::mask(crop, EOO.poly.crop[x, ])
tmp  <- raster::getValues(mask) # much faster than raster::extract
head(tmp)
class(tmp)
class(tmp) %in% c("numeric","integer")
tmp <- tmp[!is.na(tmp)] # excluding pixels outside the EOO
head(tmp)
#Getting the approximated area of the masked raster (in km2)
mask.area <- raster::area(mask[[1]], na.rm = TRUE)
mask
image(mask)
raster::image(mask)
proj_crs <- proj_crs(proj_type = proj_type)
proj_crs
proj_hab <- raster::crs(hab.map)
proj_hab
is.na(sf::st_crs(EOO.poly))
EOO.poly <- sf::st_transform(EOO.poly, crs = proj_hab)
EOO.poly
image(hab.map)
raster::image(hab.map)
plot(EOO.comp[,1], add=TRUE
)
plot(EOO.poly[,1], add=TRUE
)
raster::image(hab.map)
plot(EOO.poly[,1], add=TRUE
)
proj_crs <- proj_crs(proj_type = proj_type)
proj_hab <- raster::crs(hab.map)
if(is.na(sf::st_crs(EOO.poly)))
sf::st_crs(EOO.poly) <- 4326
EOO.poly <- sf::st_transform(EOO.poly, crs = proj_hab)
# Croping EOO shapefiles to the raster extent
EOO.poly.crop <- suppressWarnings(
sf::st_intersection(EOO.poly, sf::st_as_sfc(sf::st_bbox(hab.map))))
if (length(EOO.poly$geometry) > length(EOO.poly.crop$geometry))
warning(paste0("The EOO of ",
length(EOO.poly$geometry) - length(EOO.poly.crop$geometry),
" species became empty after croping them to the habitat map extent"))
EOO.poly.proj <- sf::st_transform(EOO.poly.crop, crs = proj_crs)
EOO.poly.area <- as.double(sf::st_area(EOO.poly.proj))/1000000 # area in km2
EOO.poly.name <- EOO.poly.crop
sf::st_geometry(EOO.poly.name) <- NULL
EOO.poly.name <- EOO.poly.name[,1]
raster::image(hab.map)
plot(EOO.poly.crop[,1], add=TRUE
)
plot(EOO.poly[,1]
)
raster::image(hab.map, add=TRUE)
raster::image(hab.map)
plot(EOO.poly[,1]
)
raster::image(hab.map, add=TRUE)
raster::image(hab.map)
plot(EOO.poly[,1], add=TRUE
)
raster::image(hab.map, add=TRUE)
if(any(grepl("SpatialPolygon", class(hab.map))))
hab.map <- sf::st_as_sf(hab.map)
if (is.null(hab.class) & any(grepl("Raster", class(hab.map))))
classes <- seq(hab.map[[1]]@data@min, hab.map[[1]]@data@max, 1)
classes
length(classes)
x
ext <- raster::extent(sf::st_bbox(EOO.poly.crop[x, ])[c(1,3,2,4)])
crop <- raster::crop(hab.map, ext) # cropping raster to the extent of the polygon
mask <- raster::mask(crop, EOO.poly.crop[x, ])
tmp  <- raster::getValues(mask) # much faster than raster::extract
if(class(tmp) %in% c("numeric","integer"))
tmp <- tmp[!is.na(tmp)] # excluding pixels outside the EOO
if(class(tmp) %in% c("matrix","data.frame"))
tmp  <- tmp[!is.na(tmp[,1]),] # excluding pixels outside the EOO
crop
mask
#Getting the approximated area of the masked raster (in km2)
mask.area <- raster::area(mask[[1]], na.rm = TRUE)
mask.area
r.area <- raster::cellStats(mask.area, 'sum')
r.area
!is.null(hab.class)
class(tmp)
if(class(tmp) %in% c("integer","numeric")) {
vals <- tmp # excluding pixels outside the EOO
pixs <- length(vals)
}
if(class(tmp) %in% c("matrix","data.frame")) {
last <- dim(tmp)[2]
vals <- tmp[, last] # excluding pixels outside the EOO
pixs <- length(vals)
}
sumario <- matrix(unclass(summary(vals)), nrow = 1, ncol = 6,
dimnames = list("", c("Min","1st_Qu","Median","Mean","3rd_Qu","Max")))
sumario
plot(ext, add=TRUE, border = 5)
plot(EOO.poly.crop[x, ], add=TRUE, border = 5)
if(output_raster %in% "summary") {
res <- cbind.data.frame(tax = EOO.poly.name[x],
sumario, # ci,
stringsAsFactors = FALSE)
} else {
#Tabulating pixels
tmp1 <- table(factor(vals, levels = classes))
prop <-  matrix(round(100*unclass(tmp1)/pixs, 8), nrow = 1, ncol = length(classes),
dimnames = list("", classes))
if(output_raster %in% "prop.table")
res <- cbind.data.frame(tax = EOO.poly.name[x],
prop,
stringsAsFactors = FALSE)
if(output_raster %in% "area.table")
area <- round((prop * r.area)/100, 3)
res <- cbind.data.frame(tax = EOO.poly.name[x],
area,
stringsAsFactors = FALSE)
}
res
rm(hab.map)
rm(list=ls())
#' last rasters are used to calculate habitat loss.
#'
#'
#' @examples #To be included
#'
#' @importFrom raster crs crop extract mask
#' @importFrom sf st_crs st_transform st_intersection st_intersects st_as_sf st_geometry
#'
#'
#'
EOO.habitat <- function(EOO.poly,
hab.map = NULL,
hab.class = NULL,
years = NULL,
export_shp = FALSE,
plot = FALSE,
output_raster = "summary",
proj_type = "cea",
parallel = FALSE,
NbeCores = 2,
show_progress = TRUE
) {
proj_crs <- proj_crs(proj_type = proj_type)
proj_hab <- raster::crs(hab.map)
if(is.na(sf::st_crs(EOO.poly)))
sf::st_crs(EOO.poly) <- 4326
EOO.poly <- sf::st_transform(EOO.poly, crs = proj_hab)
# Croping EOO shapefiles to the raster extent
EOO.poly.crop <- suppressWarnings(
sf::st_intersection(EOO.poly, sf::st_as_sfc(sf::st_bbox(hab.map))))
if (length(EOO.poly$geometry) > length(EOO.poly.crop$geometry))
warning(paste0("The EOO of ",
length(EOO.poly$geometry) - length(EOO.poly.crop$geometry),
" species became empty after croping them to the habitat map extent"))
EOO.poly.proj <- sf::st_transform(EOO.poly.crop, crs = proj_crs)
EOO.poly.area <- as.double(sf::st_area(EOO.poly.proj))/1000000 # area in km2
EOO.poly.name <- EOO.poly.crop
sf::st_geometry(EOO.poly.name) <- NULL
EOO.poly.name <- EOO.poly.name[,1]
if(any(grepl("SpatialPolygon", class(hab.map))))
hab.map <- sf::st_as_sf(hab.map)
if (is.null(hab.class) & any(grepl("Raster", class(hab.map))))
classes <- seq(hab.map[[1]]@data@min, hab.map[[1]]@data@max, 1)
# Extracting the map information for each EOO polygon
if (parallel) {
cl <- snow::makeSOCKcluster(NbeCores)
doSNOW::registerDoSNOW(cl)
message('Parallel running with ',
NbeCores, ' cores')
`%d%` <- foreach::`%dopar%`
} else{
`%d%` <- foreach::`%do%`
}
x <- NULL
if (show_progress) {
pb <-
txtProgressBar(
min = 0,
max = length(EOO.poly.crop$geometry),
style = 3
)
progress <- function(n)
setTxtProgressBar(pb, n)
opts <- list(progress = progress)
} else{
opts <- NULL
}
output <-
foreach::foreach(
x = 1:length(EOO.poly.crop$geometry),
#x = 4000:5000,
#.combine = 'c',
#.packages=c("raster","sf"),
.options.snow = opts
) %d% {
# if (!parallel & show_progress)
#   setTxtProgressBar(pb, x)
if(any(grepl("sf", class(hab.map)))) {
crop <- sf::st_intersection(hab.map, EOO.poly.crop[x, ]) # cropping raster to the extent of the polygon
crop.proj <- sf::st_transform(crop, crs = proj_crs)
area.hab <- as.numeric(sum(sf::st_area(crop), na.rm = TRUE))/1000000
tmp <- c(as.numeric(sf::st_area(EOO.poly.crop[x, ]))/1000000 - area.hab, area.hab)
tmp1 <- crop
sf::st_geometry(tmp1) <- NULL
n.polys <- length(unique(as.character(tmp1[,1]))) ### CHECK HERE: FAIL TO GET THE NUMBER OF SHAPE IDs
hab.mat <- matrix(c(NA,  n.polys, 100*tmp/EOO.poly.area[x], tmp), ncol = 3, nrow = 2, byrow = FALSE)
row.names(hab.mat) <- c("non_habitat", "habitat")
colnames(hab.mat) <- c("numb.polys","prop.EOO", "area.EOO")
res <- cbind.data.frame(tax = EOO.poly.name[x], hab.mat,
stringsAsFactors = FALSE)["habitat",]
if(export_shp) {
## Tried to combine the two sf objects but could not make it with the df info
# toto <- c(st_geometry(crop), st_geometry(EOO.poly.crop[x, ]))
# merge(toto, crop)
# toto <- sf::st_multipolygon(st_geometry(eoo1), st_geometry(crop1))
# crop1 <- st_cast(crop, "POLYGON")
# eoo1 <- st_cast(EOO.poly.crop[x,], "POLYGON")
res <- list(res, crop)
if (plot) {
par(mar = c(3,3,2,2), las=1, tcl=-0.25, mgp=c(2.5,0.5,0))
plot(sf::st_geometry(EOO.poly.crop[x, ]),
main = EOO.poly.name[x], bg=0)
plot(sf::st_geometry(crop), add = TRUE, col = 1)
}
}
}
# result <- NULL
# for(x in 4000:5000) {
if(grepl("Raster", class(hab.map))) {
ext <- raster::extent(sf::st_bbox(EOO.poly.crop[x, ])[c(1,3,2,4)])
crop <- raster::crop(hab.map, ext) # cropping raster to the extent of the polygon
mask <- raster::mask(crop, EOO.poly.crop[x, ])
tmp  <- raster::getValues(mask) # much faster than raster::extract
if(class(tmp) %in% c("numeric","integer"))
tmp <- tmp[!is.na(tmp)] # excluding pixels outside the EOO
if(class(tmp) %in% c("matrix","data.frame"))
tmp  <- tmp[!is.na(tmp[,1]),] # excluding pixels outside the EOO
#Getting the approximated area of the masked raster (in km2)
mask.area <- raster::area(mask[[1]], na.rm = TRUE)
r.area <- raster::cellStats(mask.area, 'sum')
if (!is.null(hab.class)) {
# Getting habitat quantity (land-use maps)
mask.dim <- dim(tmp)
pixs <- mask.dim[1]
last <- mask.dim[2]
hab.mat <- as.matrix(table(factor(tmp[, last] %in% hab.class, levels = c(FALSE, TRUE))))
row.names(hab.mat) <- c("non_habitat", "habitat")
hab.mat <- cbind(hab.mat, 100 * hab.mat/pixs)
hab.mat <- cbind(hab.mat, (hab.mat[,2]* r.area)/100)
colnames(hab.mat) <- c("n.pixs", "prop.EOO", "area.EOO")
if (class(crop) == "RasterBrick") {
# Getting the overal percentage of each transition
transit.mat <- 100 * as.matrix(table(factor(tmp[, 1] %in% hab.class, levels = c(FALSE, TRUE)),
factor(tmp[, last] %in% hab.class, levels = c(FALSE, TRUE)))) / pixs
colnames(transit.mat) <- c("non_habitat_t1", "habitat_t1")
row.names(transit.mat) <- c("non_habitat_t0", "habitat_t0")
hab.mat <- cbind(hab.mat, loss = c(transit.mat[1, 2], transit.mat[2, 1]))
hab.mat <- cbind(hab.mat, recover = c(transit.mat[2, 1], transit.mat[1, 2]))
# % of habitat loss and recovery
hab_loss <- 100 * transit.mat[2, 1] / sum(transit.mat[2, ])
non_hab_loss <- 100 * transit.mat[1, 2] / sum(transit.mat[1, ])
hab_rec <- 100 * transit.mat[1, 2] / sum(transit.mat[, 2])
non_hab_rec <- 100 * transit.mat[2, 1] / sum(transit.mat[, 1])
hab.mat <- cbind(hab.mat, rel.loss = c(non_hab_loss, hab_loss))
hab.mat <- cbind(hab.mat, rel.recover = c(non_hab_rec, hab_rec))
# rate of loss and recovery
if(is.null(years))
years <- dim(hab.map)[3] - 1
non_hab_loss_rate <- non_hab_loss / years
hab_loss_rate <- hab_loss / years
non_hab_rec_rate <- non_hab_rec / years
hab_rec_rate <- hab_rec / years
hab.mat <- cbind(hab.mat, rate.loss = c(non_hab_loss_rate, hab_loss_rate))
hab.mat <- cbind(hab.mat, rate.recover = c(non_hab_rec_rate, hab_rec_rate))
# defining habitat quality at time t+1
transit <- matrix(NA, ncol = 1, nrow = pixs)
transit[tmp[, 1] %in% hab.class &
tmp[, last] %in% hab.class, 1] <- 2
transit[tmp[, 1] %in% hab.class &
!tmp[, last] %in% hab.class, 1] <- -2
transit[!tmp[, 1] %in% hab.class &
tmp[, last] %in% hab.class, 1] <- 1
transit[!tmp[, 1] %in% hab.class &
!tmp[, last] %in% hab.class, 1] <- 0
mod <- stats::lm(base::jitter(transit, factor=0.1) ~ 1)
ci <- round(stats::confint(mod),4)
hab.mat <- cbind(hab.mat,
hab.quality = round(as.numeric(stats::coef(mod)),4),
hab.quality.lo = ci[1],
hab.quality.hi = ci[2])
res <- cbind.data.frame(tax = EOO.poly.name[x], hab.mat[,-1],
stringsAsFactors = FALSE)["habitat",]
# location of decline and recover
if (export_shp) {
#mask <- raster::mask(crop, EOO.poly.crop[x,], cellnumbers = TRUE)
loc.loss <- data.frame(
cell = 1:length(mask[[1]]@data@values),
raster::coordinates(mask[[1]]),
classes = NA_integer_
)
#loc.loss$classes[loc.loss$cell %in% tmp[, 1]] <- transit
loc.loss$classes[!is.na(mask[[1]]@data@values)] <- transit
r <- raster::rasterFromXYZ(loc.loss[!is.na(loc.loss$classes), 2:4],
crs = raster::crs(mask[[1]]))
if (plot) {
par(mar = c(3,3,2,2), las=1, tcl=-0.25, mgp=c(2.5,0.5,0))
raster::plot(r, col = c("red", "grey", "green", "darkgreen"),
main = EOO.poly.name[x])
plot(sf::st_geometry(EOO.poly.crop[x,]), add = TRUE, bg = 0)
}
res <- list(res, r)
}
}
} else {
#Getting summary stats for the variable in the EOO (quantitative maps)
if(class(tmp) %in% c("integer","numeric")) {
vals <- tmp # excluding pixels outside the EOO
pixs <- length(vals)
}
if(class(tmp) %in% c("matrix","data.frame")) {
last <- dim(tmp)[2]
vals <- tmp[, last] # excluding pixels outside the EOO
pixs <- length(vals)
}
sumario <- matrix(unclass(summary(vals)), nrow = 1, ncol = 6,
dimnames = list("", c("Min","1st_Qu","Median","Mean","3rd_Qu","Max")))
# if(class(vals) %in% "integer") {
#   mod <- stats::glm(vals ~ 1, family = "poisson" )
#   ci <- suppressMessages(round(exp(stats::confint(mod)),4))
# }
# if(class(tmp) %in% "numeric") {
#   mod <- stats::lm(vals ~ 1)
#   ci <- suppressMessages(round(stats::confint(mod),4))
# }
if(output_raster %in% "summary") {
res <- cbind.data.frame(tax = EOO.poly.name[x],
sumario, # ci,
stringsAsFactors = FALSE)
} else {
#Tabulating pixels
tmp1 <- table(factor(vals, levels = classes))
prop <-  matrix(round(100*unclass(tmp1)/pixs, 8), nrow = 1, ncol = length(classes),
dimnames = list("", classes))
if(output_raster %in% "prop.table")
res <- cbind.data.frame(tax = EOO.poly.name[x],
prop,
stringsAsFactors = FALSE)
if(output_raster %in% "area.table")
area <- round((prop * r.area)/100, 3)
res <- cbind.data.frame(tax = EOO.poly.name[x],
area,
stringsAsFactors = FALSE)
}
}
}
res
# if(is.null(result)) result <- res else result <- rbind.data.frame(result, res, stringsAsFactors = FALSE)
# cat(x, "\n")
}
if(parallel) snow::stopCluster(cl)
if(show_progress) close(pb)
result <- do.call(rbind.data.frame, output)
rownames(result) <- NULL
return(result)
}
### HUMAN INFLUENCES ### between 1995 - 2004 (0 low and 65 is high)
hab.map <- raster::raster(paste(path,"//Am_Lat_Global_Human_Influence_Index_v2//hii_s_amer//w001001.adf",sep=""))
## ESA land cover
path <- "E://ownCloud//W_GIS" # Renato's path
### HUMAN INFLUENCES ### between 1995 - 2004 (0 low and 65 is high)
hab.map <- raster::raster(paste(path,"//Am_Lat_Global_Human_Influence_Index_v2//hii_s_amer//w001001.adf",sep=""))
hab.map
hii <- raster::raster(paste(path,"//Am_Lat_Global_Human_Influence_Index_v2//hii_global//hii_neotrop_meth3.tif",sep=""))
hii
### HUMAN INFLUENCES ### between 1995 - 2004 (0 low and 65 is high)
hab.map <- raster::raster(paste(path,"//Am_Lat_Global_Human_Influence_Index_v2//hii_neotrop.tif",sep=""))
EOO.habitat(EOO.poly[1:50,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 4)
## EOO shapefiles
EOO.poly <- readRDS(file = "C:/Users/renato/Documents/raflima/Pos Doc/Manuscritos/Artigo Extincao na MA/data analysis/data/spp.convex.hull.polys_sf_uncropped.rds")
EOO.habitat(EOO.poly[1:50,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 4)
t0 <- Sys.time()
toto2.1 <- EOO.habitat(EOO.poly[1:1000,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 6)
toto2.2 <- EOO.habitat(EOO.poly[1001:2000,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 6)
toto2.3 <- EOO.habitat(EOO.poly[2001:3000,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 6)
toto2.4 <- EOO.habitat(EOO.poly[3001:4000,], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 6)
toto2.5 <- EOO.habitat(EOO.poly[4001:length(EOO.poly$geometry),], hab.map, hab.class = NULL, years = NULL,
parallel = TRUE, NbeCores = 6)
toto2 <- cbind.data.frame(toto2.1, toto2.2, toto2.3, toto2.4, toto2.5,
stringsAsFactors = FALSE)
Sys.time() - t0
dim(toto2.1)
dim(toto2.2)
dim(toto2.3)
dim(toto2.4)
dim(toto2.5)
toto2 <- rbind.data.frame(toto2.1, toto2.2, toto2.3, toto2.4, toto2.5,
stringsAsFactors = FALSE)
saveRDS(toto2, "C:/Users/renato/Documents/raflima/Pos Doc/Manuscritos/Artigo Extincao na MA/data analysis/data/EOO_HII.rds")
## ESA land cover
path <- "E://ownCloud//W_GIS" # Renato's path
fcov <- raster::stack(paste(path,"//WO_ESA_Land_Cover_map//v2.0.7//ESACCI-LC-L4-LCCS-Map-300m-P1Y-1992_2015-v2.0.7_AF_50km_buffer_mask.tif",sep=""))
names(fcov) = paste("ESA",c(1992,1995,2000,2005,2010,2015), sep=".")
hab.map <- fcov[[c(1,6)]]
toto = read.csv(paste(path,"//WO_ESA_Land_Cover_map//v2.0.7//ESACCI-LC-Legend.csv",sep=""), as.is=TRUE)
hab.class = toto$NB_LAB[grepl("ForestCover", toto$LegendTreeCoSimp)]
toto.1 <- EOO.habitat(EOO.poly[1:10,], hab.map, hab.class = hab.class, years = 23,
parallel = TRUE, NbeCores = 6)
toto.1