Dawn2: Executive function for testing AOA

In SchoenbergA/IKARUS: Classification Tool

Description

wrapper for IKARUS and AOA with buffered TrainPoints.

Usage

Dawn2(
  FFS = FALSE,
  Tpoints,
  buf_size,
  design,
  Stk,
  Stk_name,
  plot_res = TRUE,
  save_png = FALSE,
  save_res = FALSE,
  path_res,
  path_png,
  fsize = 24,
  rsp = NULL,
  rsp_class = NULL,
  validate = F
)

Arguments

FFS	boolean - if True uses a FFS. Default= FALSE
Tpoints	SpatialPoints - PointLayer with the TrainingPositions
buf_size	numeric - with for the buffer in meter
design	character - from for the buffer. Choose from "ROUND","FLAT","SQUARE"
Stk	RasterStack - with predictors for classification
Stk_name	character - name of input Stack for the name to write in images and output name.
plot_res	boolean - set 'TRUE' to plot the prediction and AOA. Default = TRUE.
save_png	boolean - set 'TRUE' to write the images in.png format for prediction and AOA. Default = FALSE.
save_res	boolean - set 'TRUE' to write the prediction and AOA in .grd format. Default = FALSE.
path_res	character - the path to save the resulting RasterLayer. Required if "save_res" option is used.
fsize	numeric - the desired front size for the main and sub text. default = 24.
rsp	SpatialPolygons - Response Polygon for a class used for validation.
rsp_class	numeric - The position of the response class in the
path_png_cl	character - the path to save the resulting plots for classification as png. Required if "save_png" option is used.
path_png_aoa	character - the path to save the resulting plots for AOA png. Required if "save_png" option is used.

Details

This function is a wrapper for the IKARUS workflow for a LLOCV Random Forest classification and further uses the AOA approach by Meyer 2020. By default the resulting prediction and AOA are plotted. By default further uses all varibales for the model, optional uses a FFS (if FFS=TRUE). Optional saves the RasterLayers. Further can save the results to .png format to a desired path. Further the 'rsp' argument allow the direct estimation of the model performance by comparing one selected class with a response layer.

Value

Returns the classification, model and the AOA. Optional saves the resulting RasterLayer and saves resuöting prediction and AOA to .png images.

Author(s)

Andreas Schönberg

SchoenbergA/IKARUS documentation built on Sept. 8, 2021, 11:11 a.m.