Struct BlockCrossValidator 

pub struct BlockCrossValidator { /* private fields */ }

Block cross-validation for time series or sequential data

Splits the data into blocks where each fold uses a contiguous block for testing and the preceding data for training. This is useful for time series data where we want to respect temporal order.

Example

use sklears_model_selection::{BlockCrossValidator, CrossValidator};

let block_cv = BlockCrossValidator::new(3);
let splits = block_cv.split(12, None);
// This would create 3 folds with blocks of 4 samples each

Implementations

impl BlockCrossValidator

pub fn new(n_splits: usize) -> Self

Create a new block cross-validator

Arguments

• n_splits - Number of blocks to create

pub fn test_size(self, test_size: usize) -> Self

Set the size of each test block

pub fn gap(self, gap: usize) -> Self

Set the gap between train and test sets to avoid data leakage

Trait Implementations

impl Clone for BlockCrossValidator

fn clone(&self) -> BlockCrossValidator

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl CrossValidator for BlockCrossValidator

fn n_splits(&self) -> usize

Returns the number of splits

fn split( &self, n_samples: usize, _y: Option<&Array1<i32>>, ) -> Vec<(Vec<usize>, Vec<usize>)>

Generate train/test indices for cross-validation

impl Debug for BlockCrossValidator

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

Formats the value using the given formatter.