In jscamac/edmaps: Estimate Likelihood of Pest Establishment

Estimating establishment likelihood for species

edmaps is best used in reproducible workflows powered by drake, with each step explicitly represented within a plan. Plans specify R expressions (commands) and targets that will store the commands' returned value. The process of estimating establishment likelihood begins with the preparation of core datasets that are used across species. Subsequent targets relate to individual species to be assessed.

A pair of functions has been included to facilitate generation of these drake plans.

• species_plan(): Given a set of user inputs, the output of species_plan is a new drake plan that contains the targets required to estimate risk for the species, as well as the preliminary data wrangling steps.
• excel_to_plan(): This function provides a simple way to generate plans for species, and is particularly useful for generating plans for multiple species as a batch. It is used by providing the path to a Microsoft Excel xlsx file containing parameter values for one or more species (example at the path given by system.file('extdata/parameters.xlsx', package='edmaps')), in which each row of the xlsx corresponds to a species, and columns contain parameters relating to the species. When called, excel_to_plan imports and parses the xlsx file, and then iterates over species, passing the appropriate arguments to species_plan and combining the resulting plans.

Plans generated by the above processes (as demonstrated below) can then be built using drake::make(plan).

Creating a plan for a single species

A plan defining the steps involved in risk estimation for Bactrocera dorsalis can be created using species_plan as follows:

library(edmaps)

dorsalis <- species_plan(
  species='Bactrocera dorsalis',
  aggregated_res=c(10000, 10000),
  clum_path='data/biotic/raw_data/ACLUM/clum_50m1218m.tif',
  nvis_path='data/edmaps/biotic/raw_data/NVIS_5.1/aus5_1e_mvs',
  ndvi_path='data/edmaps/biotic/raw_data/NDVI/NDVI_Oct18_Mar19.grid',
  fertiliser_data_path='data/edmaps/pathway/raw_data/Fertiliser/fertiliser2016_17.csv',
  nrm_path='data/edmaps/pathway/raw_data/Fertiliser/nrm_regions/NRMR_2016_AUST.shp',
  containers_data_path='data/edmaps/pathway/raw_data/Containers/containers_bypostcode.xls',
  postcode_path='data/edmaps/pathway/raw_data/Containers/postal_areas/POA_2011_AUST.shp', 
  airports_path='data/edmaps/pathway/raw_data/Major_Airports/MajorAviationTerminals.gdb',
  tourist_beds_path='data/edmaps/pathway/raw_data/Tourist_Beds2016/TourBedsAlbers1km.shp',
  pop_density_path='data/edmaps/pathway/raw_data/Population2016/pop_density_1000m.tif',
  processed_data_path='data/processed',
  clum_classes=c(340, 341, 342, 344, 345, 347, 348, 349, 350, 351, 353, 365,
                 440, 441, 442, 444, 445, 447, 448, 449, 450, 451, 453, 540, 
                 541, 542, 543, 544),
  pathways=c('tourists', 'residents', 'torres'),
  include_abiotic_weight=TRUE,
  occurrence_path='data/abiotic/occurrences/oriental_fruitfly/dorsalis_occurrences.csv',
  cabi_path='data/abiotic/occurrences/oriental_fruitfly/cabi_dorsalis_20April19.csv',
  use_gbif=TRUE,
  manual_check_flagged_records=FALSE,
  gbif_species=c("Bactrocera dorsalis","Bactrocera invadens", 
                 "Bactrocera papayae", "Bactrocera philippinensis"),
  leakage_tourists=c(7, 67), establishment_tourists=c(0.001, 0.01),
  leakage_returning=c(3, 33), establishment_returning=c(0.001, 0.01),
  leakage_torres=c(1, 100), establishment_torres=c(0.001, 0.01))

# Let's take a look
dorsalis

# List the targets
dorsalis$target

Creating a plan for multiple species

To create a plan for multiple species, it is simpler to use excel_to_plan. Using the example xlsx file included with edmaps, we can prepare a plan for four species: oriental fruitfly, khapra beetle, brown marmorated stink bug and gypsy moth. Below, we point excel_to_plan to that file. Note that this example requires a number of raw data files that are not distributed with the edmaps package.

f <- system.file('extdata/parameters.xlsx', package='edmaps')
all <- excel_to_plan(f)

# Let's take a look
all

# List the targets
all$target

jscamac/edmaps documentation built on June 11, 2022, 1:26 a.m.