Struct KalmanFilter 

pub struct KalmanFilter { /* private fields */ }

Kalman filter with a Local Linear Trend state-space model.

Example

use quant_indicators::KalmanFilter;
use quant_primitives::Candle;
use chrono::Utc;
use rust_decimal_macros::dec;

let ts = Utc::now();
let candles: Vec<Candle> = (0..100).map(|i| {
    let price = dec!(100) + rust_decimal::Decimal::from(i);
    Candle::new(price, price, price, price, dec!(1000), ts).unwrap()
}).collect();
let kf = KalmanFilter::new(dec!(0.01), dec!(0.001), dec!(1), 20, 20).unwrap();
let result = kf.compute(&candles).unwrap();
assert_eq!(result.len(), 80); // 100 - 20

Implementations

impl KalmanFilter

pub fn new( q_level: Decimal, q_slope: Decimal, r_obs: Decimal, warmup: usize, eiv_window: usize, ) -> Result<Self, IndicatorError>

Create a new Kalman filter.

Arguments

• q_level — process noise for level state (must be > 0)
• q_slope — process noise for slope state (must be > 0)
• r_obs — measurement noise (must be > 0)
• warmup — bars consumed before output starts (must be ≥ 1)
• eiv_window — rolling window size for empirical innovation variance (must be ≥ 2)

Errors

Returns InvalidParameter if any noise parameter is ≤ 0, warmup is 0, or eiv_window is less than 2.

pub fn name(&self) -> &str

Get the indicator name.

pub fn warmup_period(&self) -> usize

Minimum number of candles consumed before output starts.

pub fn compute( &self, candles: &[Candle], ) -> Result<KalmanResult, IndicatorError>

Compute the Kalman filter on candle data.

Returns KalmanResult with five output series, each of length candles.len() - warmup.

Errors

Returns InsufficientData if candles.len() <= warmup.

Trait Implementations

impl Clone for KalmanFilter

fn clone(&self) -> KalmanFilter

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for KalmanFilter

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

Formats the value using the given formatter.