Struct RollingOriginSplit 

pub struct RollingOriginSplit {
    pub series_len: usize,
    pub min_train: usize,
    pub horizon: usize,
    pub gap: usize,
    pub step: usize,
    pub folds: Arc<[TemporalFold]>,
}

Rolling-origin (expanding window) temporal folds.

Fields

series_len: usize

Series length.

min_train: usize

Minimum initial training length.

horizon: usize

Forecast / test horizon per fold.

gap: usize

Gap between train end and test start.

step: usize

Step between successive fold origins.

folds: Arc<[TemporalFold]>

Ordered folds.

Implementations

impl RollingOriginSplit

pub fn try_new( series_len: usize, min_train: usize, horizon: usize, gap: usize, step: usize, ) -> Result<Self, DataError>

Build expanding-origin folds.

Fold k uses train [0, origin) where origin = min_train + k * step, then gap, then test of length horizon.

Errors

When no complete fold fits in series_len.

Trait Implementations

impl Clone for RollingOriginSplit

fn clone(&self) -> RollingOriginSplit

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RollingOriginSplit

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

Formats the value using the given formatter.

impl Eq for RollingOriginSplit

impl PartialEq for RollingOriginSplit

fn eq(&self, other: &RollingOriginSplit) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for RollingOriginSplit