Struct RunningStats 

pub struct RunningStats { /* private fields */ }

Per-dimension running mean and variance tracker using Welford’s algorithm.

Implementations

impl RunningStats

pub fn new(dim: usize) -> Self

Create a new tracker for vectors of dimension dim.

pub fn dim(&self) -> usize

Dimension (number of tracked features).

pub fn count(&self) -> u64

Number of samples seen so far.

pub fn mean_f32(&self) -> Vec<f32>

Per-dimension mean.

pub fn std_f32(&self) -> Vec<f32>

Per-dimension sample standard deviation (returns [1.0; dim] until count ≥ 2).

pub fn var_f32(&self) -> Vec<f32>

Per-dimension variance.

pub fn update(&mut self, obs: &[f32]) -> RlResult<()>

Update statistics with a new observation vector.

Errors

• RlError::DimensionMismatch if obs.len() != dim.

pub fn update_batch(&mut self, batch: &[f32]) -> RlResult<()>

Update statistics with a batch of observations [B × dim].

Errors

• RlError::DimensionMismatch if batch.len() % dim != 0.

pub fn normalise(&self, obs: &[f32]) -> RlResult<Vec<f32>>

Normalise a single observation: (obs - mean) / (std + eps).

Errors

• RlError::DimensionMismatch if obs.len() != dim.

pub fn reset(&mut self)

Reset all statistics to zero.

Trait Implementations

impl Clone for RunningStats

fn clone(&self) -> RunningStats

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RunningStats

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

Formats the value using the given formatter.