Struct StreamingRBFSampler 

pub struct StreamingRBFSampler { /* private fields */ }

Streaming RBF kernel approximation using Random Fourier Features

Maintains an online approximation of RBF kernel features that adapts to data streams with concept drift and memory constraints.

Implementations

impl StreamingRBFSampler

pub fn new(n_components: usize, gamma: f64) -> Self

Create a new streaming RBF sampler

pub fn with_config(self, config: StreamingConfig) -> Self

Set the streaming configuration

pub fn with_random_state(self, random_state: u64) -> Self

Set random state for reproducibility

pub fn fit(&mut self, x: &Array2<f64>) -> Result<()>

Initialize the streaming sampler with initial data

pub fn add_sample(&mut self, sample: StreamingSample) -> Result<()>

Add a new sample to the stream

pub fn transform(&self, x: &Array2<f64>) -> Result<Array2<f64>>

Transform data using current model

pub fn transform_sample(&self, sample: &Array1<f64>) -> Result<Array1<f64>>

Transform a single sample

pub fn buffer_stats(&self) -> (usize, f64, f64)

Get current buffer statistics

pub fn feature_stats(&self) -> &FeatureStatistics

Get feature statistics

pub fn detect_drift(&mut self, x: &Array2<f64>) -> Result<bool>

Check for concept drift