Struct ArimaModel 

pub struct ArimaModel { /* private fields */ }

ARIMA(p,d,q) — non-seasonal, estimated by Burg + conditional sum of squares.

Implementations

impl ArimaModel

pub fn new(p: usize, d: usize, q: usize) -> Self

Creates a new ARIMA model with the given orders.

pub fn aic(&self) -> Option<f64>

Akaike Information Criterion, n·ln(RSS/n) + 2k.

n counts only the innovations the conditional likelihood is defined over, which is what makes the value comparable between two models of the same series at the same differencing order. It is not comparable across differencing orders — a differenced series is a different series — which is why auto_arima chooses d by a unit-root test rather than by this number.

pub fn bic(&self) -> Option<f64>

Bayesian Information Criterion, n·ln(RSS/n) + k·ln(n).

pub fn aicc(&self) -> Option<f64>

Corrected Akaike Information Criterion (AICc).

pub fn diagnostics( &self, actual: &[f64], predicted: &[f64], ) -> Option<ModelDiagnostics>

Comprehensive model diagnostics.

Requires both the original training values and the in-sample predictions (fitted values) for the model.

Trait Implementations

impl<'de> Deserialize<'de> for ArimaModel

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl ForecastModel for ArimaModel

fn fit( &mut self, timestamps: &[i64], values: &[f64], ) -> Result<(), ForecastError>

Fit the model on historical data.

fn predict(&self, horizon: usize) -> Result<ForecastResult, ForecastError>

Predict horizon steps ahead.

fn update(&mut self, timestamp: i64, value: f64) -> Result<(), ForecastError>

Online update with a single new observation.

fn model_type(&self) -> ModelType

Returns the model type.

fn params(&self) -> &ModelParams

Returns the fitted parameters.

impl ModelStore for ArimaModel

fn to_bytes(&self) -> Result<Vec<u8>, StorageError>

Serializes this model to a byte vector.

fn from_bytes(bytes: &[u8]) -> Result<Self, StorageError>

where Self: Sized,

Deserializes a model from bytes.

impl Serialize for ArimaModel

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.