Skip to contents

Using raster file

Source: vignettes/articles/Using-raster-file.Rmd
Using-raster-file.Rmd

This is a basic example which shows you how to run sensitivity analysis on pre-loaded parameters.As a pre-requisite, install package using

if (!require("geohabnet")) {
  devtools::install_github("GarrettLab/CroplandConnectivity", subdir = "geohabnet")
}

Package offers 2 ways to use custom raster file - using parameters.yaml and sa_onrasters which is also used internally to create raster object using the provided source.

Loading package

if (!require(geohabnet)) {
  install.packages("geohabnet")
  library(geohabnet)
}
#> Loading required package: geohabnet

Set the file parameter under Hosts to the absolute path of raster file -

Hosts:
      monfreda: ["avocado"]
      mapspam:
      file: "full-path-to-tif/avocado_HarvestedAreaFraction.tif"

Feed the updated parameters to package -

set_parameters("parameters.yaml")

Raster files ends with extension .tif. Since, the internal workflow only uses harvested fractional area of a cropland, we recommend using raster files as specified in this dataset. For this particular example, we will use spam2017V2r1_SSA_H_BEAN_A.tif from the monfreda.

# Generate raster object
# replace the tif file with actual path after download
rr <- get_crop_raster_fromtif("spam2017V2r1_SSA_H_BEAN_A.tif")
rr
class      : RasterLayer
dimensions : 2160, 4320, 9331200  (nrow, ncol, ncell)
resolution : 0.083333, 0.083333  (x, y)
extent     : -180, 179.9986, -89.99928, 90  (xmin, xmax, ymin, ymax)
crs        : +proj=longlat +datum=WGS84 +no_defs
source     : spam2017V2r1_SSA_H_BEAN_A.tif
names      : spam2017V2r1_SSA_H_BEAN_A

To ensure we have the correct raster,

terra::plot(rr)
Raster plot for Beans
Raster plot for Beans

Next, once you have all the values ready, run risk analysis with geohabnet::sa_onrasters()

geohabnet::sa_onrasters(cropharvest_raster = rr,
                        geo_scales = list(c(-50, 50, -50, 50)),
                        link_thresholds = c(0),
                        host_density_thresholds = c(0.0000015),
                        res = 24)

On successful analysis, function will produce output maps. See example below.

Difference map
Difference map

If the extent is incorrect and no valid cells can be discovered based threshold, then analysis will be interrupted and error will be thrown - 

Error in .extract_cropland_density(the$cropharvest_aggtm_crop, host_density_threshold) : 
  host density threshold:  0.00015  is greater than the max value:  -Inf  of aggregate raster