R/bootstrap.residuals.R

In nikosbosse/SAE: This Package implements a Small Area Estimation approach from Elbers. et al blabla

# This is the loop for the residuals, it can be implemented in two different ways,
# either the first, completely disregarding the location drawn
# (F: I think this is oversimplifying it too much) and the second where depending on the location
# drawn a random residual from that cluster is drawn



# Alternative 1.3
bootstrap.residuals <-
  function(n_boot1,
           sample_size,
           num_residuals,
           residuals) {
    return(rddrawIndicesC3(num_residuals, sample_size * n_boot1, residuals))
  }




# Alternative 1.0
bootstrap.residuals1.0 <- function(n_boot1, sample_size, num_residuals, residuals){

  bootstap_indices <- sample(x = 1:num_residuals, size = sample_size * n_boot1, replace = T)

  matrix(residuals[bootstap_indices], ncol = n_boot1)
}
# This bootstrap function just randomly draws from all residuals and completely ignores the locations

#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
#
# Rcpp::cppFunction('
#   NumericVector sample(int lower, int upper, int samplelength) {
#     int i;  // index for elements in stream[]
#
#     // reservoir[] is the output array. Initialize it with
#     // first k elements from stream[]
#     int reservoir[k];
#                   for (i = 0; i < k; i++)
#                   reservoir[i] = stream[i];
#
#                   // Use a different seed value so that we don't get
#                   // same result each time we run this program
#                   srand(time(NULL));
#
#                   // Iterate from the (k+1)th element to nth element
#                   for (; i < n; i++)
#                   {
#                     // Pick a random index from 0 to i.
#                     int j = rand() % (i+1);
#
#                     // If the randomly  picked index is smaller than k, then replace
#                     // the element present at the index with new element from stream
#                     if (j < k)
#                       reservoir[j] = stream[i];
#
#   }
# ')
#
#
#
# Rcpp::cppFunction()
#
#
# #
# #
# #
# # ### complicated bootstrap formula. We could also use the easy one.
# # # this one draws residuals that correspond to the random location effect that was drawn previously
# #
# #
# # bootstrap.residuals.compl <- function(n_boot = n_boot,
# #                                       location_and_obs = num_obs_in_location,
# #                                       bootstrap_location = random_location_boot){
# #
# #   #create a dataframe where we know how often each region got bootstrapped
# #   num_obs_bootstrap <- num.obs.in.location(bootstrap_location$location)
# #
# #   #draw indices. This assumes that the original data is ordered according to locations.
# #   # N: (we should do that somewhere)
# #   n_locations <- nrow(location_and_obs)
# #
# #   bootstrap_res <- rep(NA, times = nrow(random_location_boot))
# #   counter <- 0
# #   for (i in 1:n_locations){
# #
# #     bootstrap_res[seq.int(from = i, to = i + num_obs_bootstrap[i,2])] <-
# #       sample(x = seq.int(from = i, to = i + num_obs_in_location[i,2]),
# #              size = num_obs_bootstrap[i,2], replace = T) + counter
# #
# #     counter <- counter + num_obs_bootstrap[i,2]
# #   }
# #   # this loop does the following:
# #   # for every one of the different locations it wants to create a random samples of indices that point
# #   # to the residuals that were observed in these locations. To do so it
# #   # 1) initialises a counter = 0
# #   # 2) looks at the first location and sees how often this location was drawn as a random location effect
# #   #    in the first bootstrapping round
# #   # 3) if location 1 appeared 599 times in the bootstrap samples it will draw 599 indices that point to
# #   #    a random sample of residuals drawn from all residuals of location 1.
# #   # 4) to make sure the draws are residuals the counter is increased by the number of observations in the
# #   #    last location. Given that the data is ordered after locations this will generate indices that can
# #   #    be used for subsetting and drawing the wanted residuals.
# #   return(bootstrap_res)
# # }
# #
# #
# #
# #
# #
# #
# #
# # # old code
# # #--------- for every random location effect, draw corresponding epsilon  ----------#
# # epsilon_boot <- rep(NA, length(random_location_boot)) #initialise vector for loop.
# #
# # for(i in 1:nrow(random_location_boot)){
# #   epsilon_boot[i] <- sample(x = training$e_res[training$geo2_br ==
# #                                                  random_location_boot$geo2_br[i]],
# #                             size = 1)
# # }
# #
# #