Struct LaggedFeatures 

pub struct LaggedFeatures {
    pub lag1: Vec<f64>,
    pub lag5: Vec<f64>,
    pub lag10: Vec<f64>,
    pub lag21: Vec<f64>,
}

Creates lagged versions of a feature series at standard lags (1, 5, 10, 21).

Missing values at the beginning of a lag series are filled with f64::NAN.

Example

use fin_primitives::ml_features::LaggedFeatures;

let series = (1..=25).map(|i| i as f64).collect::<Vec<_>>();
let lagged = LaggedFeatures::compute(&series);
// lag-1 of index 1 = series[0] = 1.0
assert!((lagged.lag1[1] - 1.0).abs() < 1e-10);
// lag-5 of index 0 is NaN
assert!(lagged.lag5[0].is_nan());

Fields

lag1: Vec<f64>

Original series lagged by 1 period.

lag5: Vec<f64>

Original series lagged by 5 periods.

lag10: Vec<f64>

Original series lagged by 10 periods.

lag21: Vec<f64>

Original series lagged by 21 periods.

Implementations

impl LaggedFeatures

pub fn compute(series: &[f64]) -> Self

Compute lagged feature series from a price/return series.

The output vectors have the same length as the input. Positions before the lag are f64::NAN.

pub fn as_named_pairs(&self) -> Vec<(&'static str, &[f64])>

Returns all four lag series as a vec of (lag_name, values) pairs.

Trait Implementations

impl Clone for LaggedFeatures

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for LaggedFeatures

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

Formats the value using the given formatter.