From 138486ae3d273a1c5823b8edcdf4178b4ceddc80 Mon Sep 17 00:00:00 2001
From: Maria Voigt <maria.voigt@idiv.de>
Date: Mon, 30 Oct 2017 20:16:05 +0100
Subject: [PATCH] splitting prepare boot_grid into two scripts
---
.../prepare_boot_grid_resource_use.R | 94 +++++++++++++++++++
1 file changed, 94 insertions(+)
create mode 100644 src/validation/prepare_boot_grid_resource_use.R
diff --git a/src/validation/prepare_boot_grid_resource_use.R b/src/validation/prepare_boot_grid_resource_use.R
new file mode 100644
index 0000000..492ca6f
--- /dev/null
+++ b/src/validation/prepare_boot_grid_resource_use.R
@@ -0,0 +1,94 @@
+rm(list = ls())
+source("/homes/mv39zilo/projects/orangutan_density_distribution/src/preload/preload.R")
+source("/homes/mv39zilo/projects/orangutan_density_distribution/src/functions/generic/path.to.current.R")
+source("/homes/mv39zilo/projects/orangutan_density_distribution/src/functions/project_functions/values.to.tif.R")
+outdir <- "/homes/mv39zilo/work/Borneo/analysis/model_prep_and_running/results/future/"
+# for the future we need a raster with the grid_ids
+geography_path <- "/homes/mv39zilo/work/Borneo/analysis/model_prep_and_running/results/abundMod/testing_ae_and_absence/pipeline_results/ppln_ae75m_50-2017-02-28T18-00-52/geography_2015_2017-02-28.rds"
+geography_2015 <- readRDS(geography_path)%>%
+ mutate(grid_id = id, id = 1:nrow(.))%>%
+ dplyr::select(id, x_start, y_start, grid_id)
+
+
+#--------------#
+# resource use #
+#--------------#
+outdir <- "/homes/mv39zilo/work/Borneo/analysis/model_prep_and_running/results/resource_use/"
+
+resource_use <- raster("/homes/mv39zilo/work/Borneo/analysis/model_prep_and_running/results/resource_use/resource_grid_absence_country_category_id.tif")
+resource_use_grid <- as.data.frame(resource_use)
+names(resource_use_grid) <- "category"
+resource_use_grid$category <- as.character(resource_use_grid$category)
+resource_use_grid$population <- as.numeric(substr(resource_use_grid$category, start= 1, stop = 1))
+resource_use_grid$country <- as.numeric(substr(resource_use_grid$category, start= 2, stop = 2))
+resource_use_grid$province <- as.numeric(substr(resource_use_grid$category, start= 3, stop = 4))
+resource_use_grid$resource_use <- as.numeric(substr(resource_use_grid$category, start= 5, stop =5))
+resource_use_grid$grid_id <- as.numeric(substr(resource_use_grid$category, start= 6, stop = nchar(resource_use_grid$category[1])))
+
+summary(resource_use_grid)
+
+resource_use_grid[resource_use_grid$country == 3, "country"] <- "MYS"
+resource_use_grid[resource_use_grid$country == 6, "country"] <- "IDN"
+
+# for now not distinguishing Kalimantan
+resource_use_grid[resource_use_grid$country == 6, "province"] <- "KAL"
+resource_use_grid[resource_use_grid$province == 10, "province"] <- "SAB"
+resource_use_grid[resource_use_grid$province == 11, "province"] <- "SAW"
+
+summary(resource_use_grid)
+plot(table(resource_use_grid$resource_use))
+
+# # testing duplicates
+# test <- resource_use_grid
+# test <- dplyr::filter(test, grid_id != 0)
+# testtest <- test[duplicated(test$grid_id) & test$grid_id != 0, ]
+
+# we want a category with the pop number for all pixels with
+# forest
+# and the grid_id in order
+resource_use_grid[resource_use_grid$population == 2, "resource_use"] <- NA
+
+resource_use_grid[resource_use_grid$grid_id == 0 &
+ !is.na(resource_use_grid$resource_use), "resource_use"] <- NA
+
+resource_use_grid[resource_use_grid$resource_use == 0 &
+ !is.na(resource_use_grid$resource_use), "resource_use"] <- NA
+
+resource_use_grid <- resource_use_grid %>%
+ dplyr::select(resource_use, country, grid_id)
+
+# SORT WITH GRID ID
+resource_use_grid <- resource_use_grid %>%
+ dplyr::select( resource_use, country, grid_id) %>%
+ arrange(grid_id) %>%
+ filter(grid_id != 0)
+
+grid_ids <- resource_use_grid[resource_use_grid$grid_id != 0, "grid_id"]
+
+# DEAL WITH NON UNIQUE GRID-IDS, MAYBE PROBLEM WITH REPROJECTION OR SOMETHING
+# not_unique <- grid_ids[duplicated(grid_ids)]
+# resource_use_grid_not_unique <- resource_use_grid[resource_use_grid$grid_id %in% not_unique,]
+grid_ids_missing <- geography_2015$grid_id[!geography_2015$grid_id %in% grid_ids]
+grid_ids_missing <- as.data.frame(cbind(rep(NA, times = length(grid_ids_missing)),
+ rep(NA, times = length(grid_ids_missing)),
+ grid_ids_missing))
+names(grid_ids_missing) <- c("resource_use", "country", "grid_id")
+
+resource_use_grid <- resource_use_grid %>%
+ rbind(grid_ids_missing) %>%
+ arrange(grid_id)
+
+resource_use_grid[!is.na(resource_use_grid$resource_use) &
+ resource_use_grid$country != 0, "category"] <-
+ paste0(resource_use_grid[!is.na(resource_use_grid$resource_use)&
+ resource_use_grid$country != 0, "resource_use"],
+ "_",
+ resource_use_grid[!is.na(resource_use_grid$resource_use)&
+ resource_use_grid$country != 0, "country"])
+resource_use_grid$category <- as.factor(resource_use_grid$category)
+summary(resource_use_grid)
+unique(resource_use_grid$category)
+resource_use_grid <- dplyr::select(resource_use_grid, category,
+ resource_use, country, grid_id)
+
+write.csv(resource_use_grid, file.path(outdir, "resource_use_grid.csv" ), row.names = F)
--
GitLab