R/data_manipulation.R

In biogram: N-Gram Analysis of Biological Sequences

# # ' Prepare data for further analysis
# # '
# # ' Creates random stratified subsample of population.
# # ' 
# # ' @param data_set data set.
# # ' @param pos indices of positive data.
# # ' @param neg indices of negative data.
# # ' @param train_size a vector of length 2 - size of positive and negative training
# # ' subsample.
# # ' @param test_size a vector of length 2 - size of positive and negative test
# # ' subsample.
# # ' @return a list of a length 2 containing cases belonging to respectively train and test 
# # ' sample.
# # ' @details Check it for 0 sample size.
# # ' @export
# # ' @examples 
# # ' # data(iris)
# # ' # prepare_data(iris, 1:20, 101:120, c(5, 5), c(8, 8))
# 
# prepare_data <- function(data_set,
#                         pos, 
#                         neg, 
#                         train_size, # n pos, n neg
#                         test_size) {
#  id_pos <- data_sample(pos, train_size[1], test_size[1])
#  id_neg <- data_sample(neg, train_size[2], test_size[2])
#  train_dat <- data_set[c(id_pos[["train"]], id_neg[["train"]]), ]
#  test_dat <- data_set[c(id_pos[["test"]], id_neg[["test"]]), ]
#  list(train = train_dat,
#       test = test_dat)
# }
# 
# 
# # ' Subsample data
# # '
# # ' Creates random subsample of population using indices.
# # '
# # ' @param indices of data.
# # ' @param train_size size of training sample.
# # ' @param test_size size of test sample.
# # ' @return a list of a length 2 containing train and test sample indices.
# # ' @details Check it for 0 sample size.
# # ' @export
# # ' @examples data_sample(1:20, 5, 5)
# 
# data_sample <- function(indices, train_size, test_size) {
#  # sampled indices  
#  s_indices <- sample(indices, train_size + test_size)
#  train_ind <- s_indices[0L:train_size]
#  if (train_size != 0) {
#    test_ind <- s_indices[-c(0L:train_size)][0L:test_size]
#  } else {
#    test_ind <- s_indices[0L:test_size]
#  }
#  list(train = train_ind, test = test_ind)
# }