Random subsampling
Usage
random_subsampling(
sample_idx,
iterations = 10L,
test_size = 0.25,
keep_together = NULL,
balance_in_train = NULL
)Arguments
- sample_idx
Typically a numeric vector with sample index to be separated. A character vector with sample IDs could also be used
- iterations
An integer, the number of iterations in the random subsampling
- test_size
A number between 0 and 1. The samples to be included in the test set on each interation.
- keep_together
Either
NULLor a factor with the same length assample_idx.keep_togethercan be used to ensure that groups of samples are kept in together in all iterations (either on training or on test, but never split). A typical use case for this is when you have sample replicates and you want to keep all replicates together to prevent overoptimistic results (having one sample on the train subset and its replicate on the test subset would make the prediction easier to guess). Another use case for this is when you have a longitudinal study and you want to keep some subjects in the same train or test group, because you want to use some information in a longitudinal way (e.g. a multilevel plsda model).- balance_in_train
Either
NULLor a factor with the same length assample_idx.balance_in_traincan be used to force that on each iteration, the train partition contains the same number of samples of the given factor levels. For instance, if we have a dataset with 40 samples of class "A" and 20 samples of class "B", using atest_size = 0.25, we can force to always have 16 samples of class "A" and 16 samples of class "B" in the training subset. This is beneficial to those algorithms that require that the training groups are balanced.
Value
A list of length equal to iterations. Each element of the list is
a list with two entries (training and test) containing the sample_idx
values that will belong to each subset.
Examples
random_subsampling(1:100, iterations = 4, test_size = 0.25)
#> [[1]]
#> [[1]]$training
#> [1] 2 3 5 6 7 11 12 14 15 17 18 19 20 21 22 23 25 26 27
#> [20] 28 29 30 31 35 36 38 40 41 42 43 45 46 47 48 50 52 53 54
#> [39] 55 57 58 59 60 61 62 63 64 65 66 67 68 69 70 73 74 75 77
#> [58] 79 80 81 82 83 86 87 89 90 91 93 94 95 96 97 98 99 100
#>
#> [[1]]$test
#> [1] 1 4 8 9 10 13 16 24 32 33 34 37 39 44 49 51 56 71 72 76 78 84 85 88 92
#>
#>
#> [[2]]
#> [[2]]$training
#> [1] 1 2 3 4 5 6 8 11 12 13 14 16 18 19 20 21 22 23 24
#> [20] 25 27 28 29 30 31 32 33 34 35 36 37 39 40 41 42 43 44 45
#> [39] 46 47 48 52 53 54 55 56 57 59 60 61 62 64 67 70 72 73 74
#> [58] 75 78 79 80 81 83 84 85 87 89 91 92 93 95 96 97 98 100
#>
#> [[2]]$test
#> [1] 7 9 10 15 17 26 38 49 50 51 58 63 65 66 68 69 71 76 77 82 86 88 90 94 99
#>
#>
#> [[3]]
#> [[3]]$training
#> [1] 1 2 3 4 5 7 9 10 11 12 13 14 15 18 19 20 21 22 23
#> [20] 24 26 27 28 29 32 33 34 35 36 38 41 42 43 45 46 51 52 54
#> [39] 55 56 58 59 60 61 62 63 64 65 66 68 70 71 73 75 76 77 78
#> [58] 79 80 81 82 83 84 85 87 89 90 92 93 94 95 96 97 98 100
#>
#> [[3]]$test
#> [1] 6 8 16 17 25 30 31 37 39 40 44 47 48 49 50 53 57 67 69 72 74 86 88 91 99
#>
#>
#> [[4]]
#> [[4]]$training
#> [1] 2 3 7 10 11 12 13 14 15 17 18 19 20 21 22 23 24 25 26
#> [20] 27 28 29 30 31 33 36 37 39 41 43 44 45 47 48 50 51 52 55
#> [39] 56 57 58 62 63 65 66 67 68 69 70 72 74 75 76 77 78 79 80
#> [58] 81 82 83 84 85 86 87 88 89 91 92 93 94 96 97 98 99 100
#>
#> [[4]]$test
#> [1] 1 4 5 6 8 9 16 32 34 35 38 40 42 46 49 53 54 59 60 61 64 71 73 90 95
#>
#>
subject_id <- c("Alice", "Bob", "Charlie", "Eve")
random_subsampling(1:4, iterations = 2, test_size = 0.25, keep_together = subject_id)
#> [[1]]
#> [[1]]$training
#> [1] 1 2 3
#>
#> [[1]]$test
#> [1] 4
#>
#>
#> [[2]]
#> [[2]]$training
#> [1] 1 2 3
#>
#> [[2]]$test
#> [1] 4
#>
#>