Struct AlertConfig 

pub struct AlertConfig {
    pub alpha: f64,
    pub min_calibration: usize,
    pub max_calibration: usize,
    pub lambda: f64,
    pub mu_0: f64,
    pub sigma_0: f64,
    pub adaptive_lambda: bool,
    pub grapa_eta: f64,
    pub enable_logging: bool,
    pub hysteresis: f64,
    pub alert_cooldown: u64,
}

Configuration for conformal alert calibration.

Fields

alpha: f64

Significance level alpha. FPR is controlled at this level. Lower alpha = more conservative (fewer false alarms). Default: 0.05.

min_calibration: usize

Minimum calibration samples before using conformal threshold. Default: 10.

max_calibration: usize

Maximum calibration samples to retain. Default: 500.

lambda: f64

E-process betting fraction lambda. Default: 0.5.

mu_0: f64

Null hypothesis mean for standardized residuals (usually 0). Default: 0.0.

sigma_0: f64

Null hypothesis std for standardized residuals (usually 1). Default: 1.0.

adaptive_lambda: bool

Use adaptive lambda via GRAPA. Default: true.

grapa_eta: f64

GRAPA learning rate. Default: 0.1.

enable_logging: bool

Enable JSONL-compatible logging. Default: false.

hysteresis: f64

Hysteresis factor: require E_t > (1/alpha) * hysteresis to alert. Prevents alert flicker at boundary. Default: 1.1.

alert_cooldown: u64

Cooldown observations after alert before allowing another. Default: 5.

Trait Implementations

impl Clone for AlertConfig

fn clone(&self) -> AlertConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for AlertConfig

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

Formats the value using the given formatter.

impl Default for AlertConfig

fn default() -> Self

Returns the “default value” for a type.