# Training set independent samples

library(depstats)
library(progress)
library(fst)
library(dplyr)

sample_sizes <- c(50, 100, 200, 400)

#only generate samples for the remaining ones
sample_sizes_remain <- as.character(sample_sizes) %>% setdiff(list.dirs(path = "independent_samples", full.names = FALSE))

for (sample_size in sample_sizes_remain) {
  cat('working on sample size', sample_size, '...\n')

  n <- as.integer(sample_size)

  set.seed(10 * n)

  N <- 20000 #number of samples

  #generate independent samples
  X <- depstats::indgen(N, n)

  cat('working on dependence measures ...\n')
  Xind <- depstats::sampleapply(X, n, 1) %>% as.data.frame()

  cat('working on images ...\n')
  Ximage <- depstats::sampleapply(X, n, 0, grid = 25) %>% as.data.frame()

  cat('writing files ...\n')
  dir.create(paste0('./independent_samples/', sample_size)) # create subfolder

  # store results in files
  write.fst(Xind, paste0('./independent_samples/', sample_size, '/', 'ind.fst'))
  write.fst(Ximage, paste0('./independent_samples/', sample_size, '/', 'ind_image.fst'))
}