explore disimilarity metrics for a range of number of clusters

explore disimilarity metrics for a range of number of clusters

explore_dissimilarity(
  input_rast,
  output_dir,
  window_size,
  Kmeans_info,
  selected_bands = NULL,
  input_mask = NULL,
  nb_samples = 1000,
  nbClust_list = 50,
  min_sun = 0.25,
  pcelim = 0.02,
  dimPCoA = 3,
  nbCPU = 1,
  beta_metrics = "bray",
  verbose = TRUE
)

Arguments

input_rast

SpatRaster. raster to extract data from

output_dir

character. Path for output directory

window_size

numeric. window size for square plots

Kmeans_info

list. obtained from prepare_init_kmeans

selected_bands

numeric. bands selected from input_rast

input_mask

SpatRaster. mask corresponding to raster to extract data from

nb_samples

numeric. number of samples to be extracted

nbClust_list

numeric. number of clusters to be tested

min_sun

numeric. minimum percentage of sunlit pixels

pcelim

numeric. minimum proportion of pixels to consider spectral species

dimPCoA

numeric.

nbCPU

numeric. Number of CPUs available

beta_metrics

character. name of beta dissimilarity metrics Available options are "bray", "brayturn", "jaccard", "jaccardturn", "sorensen", "simpson_diss"

verbose

boolean. set true for messages

Value

list including spectral species distribution & BC diss matrix per plot, BetaPCO model