run_RF_workflow: Run the full workflow for RF

In courtiol/rangeRinPA: Protected Area Personnel and Ranger Numbers are Insufficient to Deliver Global Expectations

Run the full workflow for RF

Description

Run the full workflow for RF

Usage

run_RF_workflow(
  data,
  rerank = TRUE,
  Ncpu = 2,
  coef = 1,
  rep_feature_select = 1000,
  rep_finetune = 1000,
  rep_simu = 10000,
  n_trees = 10000,
  grid_type = "fine"
)

Arguments

data	the full dataset
rerank	whether or not to recompute variable importance recursively during selection (default = TRUE)
Ncpu	the number of CPU cores to be used (default = 1)
coef	the coefficient used to population rangers in unsurveyed part of surveyed countries (see fill_PA_area())
rep_feature_select	the number of replicates for the feature selection (default = 1000)
rep_finetune	the number of replicates for fine tuning (default = 1000)
rep_simu	the number of simulation replicates (default = 10000)
n_trees	the number of trees in the random forest
grid_type	The type of grid to use for finetuning the RF ("fine", the default, or "coarse")

Value

a list with all the output information

Examples

## Not run: 
  RF_small_test <- run_RF_workflow(data = data_rangers, Ncpu = 2, coef = 1,
                                   rep_feature_select = 2, rep_finetune = 2, rep_simu = 2,
                                   grid_type = "coarse", n_trees = 100)

## End(Not run)

courtiol/rangeRinPA documentation built on Sept. 29, 2022, 9:54 a.m.