Skip to contents

Create a Landscape with Gaps Pattern

Source: R/landscape_create_gaps.R
create_landscape_gaps.Rd

Generates a binary landscape with circular gaps (bare patches in vegetated ground). This is a convenience wrapper around create_landscape_spots with invert_landscape = TRUE by default, making "gaps" and "spots" semantically distinct pattern names for the same underlying algorithm.

Usage

create_landscape_gaps(
  width = 100,
  height = 100,
  n_spots = 15,
  spot_radius = 5,
  spot_radius_sd = 0,
  radius_noise_fraction = 0,
  regular_spots = FALSE
)

Arguments

width

Integer. Number of columns in the landscape (default: 100).

height

Integer. Number of rows in the landscape (default: 100).

n_spots

Integer. Number of circular spots to generate. For regular placement, this may be automatically reduced if the landscape cannot accommodate the requested number at the given `spot_radius`.

spot_radius

Numeric. Mean radius of each spot in cells. Must be positive and smaller than landscape dimensions.

spot_radius_sd

Numeric. Standard deviation for random variation in spot radius. Each spot's radius is sampled from N(spot_radius, spot_radius_sd). (default: 0 - no variation)

radius_noise_fraction

Numeric (0 to 1). Proportion of the spot radius where gradual edge noise is applied. 0 creates sharp circular edges, 1 applies probabilistic cell inclusion across the entire radius. For example, 0.2 means the outer 20 Works independently of `spot_radius_sd` (which varies the overall size, while this parameter affects edge sharpness).

regular_spots

Logical. If TRUE, spots are arranged on a hexagonal grid using k-means clustering. If FALSE, spots are placed randomly (default: FALSE).

Value

A landscape object with pattern "gaps" containing:

data

SpatRaster with binary values (0 = bare ground, 1 = vegetation)

pattern

Character string "gaps"

params

List of all input parameters used to generate the landscape

Details

The distinction between "spots" and "gaps":

  • gaps: Bare patches in vegetation matrix)

  • spots: Vegetation patches in bare ground)

Both patterns use the same algorithm; the pattern name primarily serves as a semantic label for training data organization.

See also

Other landscape creation: create_landscape(), create_landscape_bands(), create_landscape_bare(), create_landscape_clustered(), create_landscape_dense(), create_landscape_diffuse(), create_landscape_fingers(), create_landscape_labyrinth(), create_landscape_random(), create_landscape_sharp(), create_landscape_spots(), create_landscapes()

Examples

# Default gaps (vegetation patches in bare ground)
gaps_default <- create_landscape_gaps()

# More gaps with size variation
gaps_modified <- create_landscape_gaps(
  n_spots = 15,
  spot_radius = 8,
  spot_radius_sd = 2
)