Struct Adwin 

pub struct Adwin { /* private fields */ }

ADWIN drift detector for streaming data.

Example

use scirs2_transform::monitoring::adwin::Adwin;

let mut adwin = Adwin::new(0.002).expect("valid delta");

// Feed stable data
for i in 0..100 {
    adwin.add_element(1.0 + (i as f64) * 0.001).expect("add");
}

// Feed shifted data
for _ in 0..100 {
    let changed = adwin.add_element(50.0).expect("add");
    if changed {
        // Drift detected!
        break;
    }
}

Implementations

impl Adwin

pub fn new(delta: f64) -> Result<Self>

Create a new ADWIN detector.

• delta – confidence parameter in (0, 1). Smaller values make the detector less sensitive (fewer false positives, slower reaction). Typical values: 0.002 (default in MOA), 0.01, 0.05.

pub fn set_min_window_length(&mut self, min_len: usize)

Set the minimum window length before drift checks begin.

pub fn add_element(&mut self, value: f64) -> Result<bool>

Add an element to the window and check for change.

Returns true if a distribution change was detected (window was shrunk).

pub fn detected_change(&self) -> bool

Whether the last call to add_element detected a change.

pub fn current_mean(&self) -> f64

Current mean of the window.

pub fn current_length(&self) -> usize

Current number of elements in the window.

pub fn current_sum(&self) -> f64

Current sum of all elements in the window.

pub fn delta(&self) -> f64

The delta (confidence) parameter.

pub fn reset(&mut self)

Reset the detector to an empty state.

Trait Implementations

impl Clone for Adwin

fn clone(&self) -> Adwin

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Adwin

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

Formats the value using the given formatter.