create_training_data_from_polygons: Create Training Data from Polygons

In tabrasel/WetlandTools: General utility functions for TerrainWorks

Create Training Data from Polygons

Description

Given an input dataset and set of polygons indicating a class, sample points from inside and outside polygons and extract predictor values for positive and negative points.

Usage

create_training_data_from_polygons(
  polygons,
  predictors_raster,
  analysis_region,
  sample_rate = 0.5,
  region_margin = 50,
  polygon_class = "positive",
  nonpolygon_class = "negative"
)

Arguments

polygons	SpatVector of polygons indicating all locations belonging to the class you wish to predict
predictors_raster	SpatRaster with a layer for each predictor variable
analysis_region	polygon or raster indicating the extent from which points can be sampled. Only regions covered by non-NA cells will be included if analysis_region is a raster.
sample_rate	Samples per km^2
region_margin	width in meters of margin to draw around polygon edges which will not be used for sampling.
polygon_class	string to use for class attribute for points sampled from inside polygons
nonpolygon_class	string to use for class attribute for points sampled outside polygons

tabrasel/WetlandTools documentation built on Dec. 20, 2024, 8:50 a.m.