Struct ShuffleSplit 

pub struct ShuffleSplit { /* private fields */ }

Random-permutation cross-validation.

Yields n_splits independent train/test splits, each formed by permuting the samples and slicing off a test set and a training set. Unlike KFold, the splits are not guaranteed to partition the dataset — the same sample may be tested in several splits or none, and (if train_size + test_size < n) some samples may be left out of a given split entirely.

use model_selection_rs::splitters::{CvSplitter, ShuffleSplit, SubsetSize};

let ss = ShuffleSplit::new(5)
    .with_test_size(SubsetSize::Fraction(0.25))
    .with_seed(0);
let splits = ss.split(40).unwrap();
assert_eq!(splits.len(), 5);
assert_eq!(splits[0].1.len(), 10); // 25% of 40

Implementations

impl ShuffleSplit

pub fn new(n_splits: usize) -> Self

Create a ShuffleSplit with n_splits splits, a default test size of 10%, and the remainder used for training.

pub fn with_test_size(self, test_size: SubsetSize) -> Self

Set the test-subset size.

pub fn with_train_size(self, train_size: SubsetSize) -> Self

Set the train-subset size (defaults to “everything not in test”).

pub fn with_seed(self, seed: u64) -> Self

Set the base RNG seed.

Trait Implementations

impl Clone for ShuffleSplit

fn clone(&self) -> ShuffleSplit

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl CvSplitter for ShuffleSplit

fn split(&self, n_samples: usize) -> Result<Vec<(Vec<usize>, Vec<usize>)>>

Produce every (train_indices, test_indices) pair for n_samples rows.

fn n_splits(&self) -> usize

Number of splits this strategy yields.

impl Debug for ShuffleSplit

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.