mlr3 (version 0.1.0-9000)

ResamplingSubsampling: Subsampling Resampling

Description

repeats (default: 30) times repeated splits into training and test set with a ratio of ratio (default: 2/3) observations in the training set.

Usage

ResamplingSubsampling

Format

R6::R6Class inheriting from Resampling.

Fields

See Resampling.

Methods

See Resampling.

Examples

# NOT RUN {
# Create a task with 10 observations
task = mlr_tasks$get("iris")
task$filter(1:10)

# Instantiate Resampling
rss = mlr_resamplings$get("subsampling",
  param_vals = list(repeats = 2, ratio = 0.5))
rss$instantiate(task)

# Individual sets:
rss$train_set(1)
rss$test_set(1)
intersect(rss$train_set(1), rss$test_set(1))

# Internal storage:
rss$instance$train # list of index vectors
# }