Struct RunningStats 

pub struct RunningStats { /* private fields */ }

Numerically-stable online mean and variance using Welford’s algorithm.

Each call to push updates the running mean and the running sum-of-squared-deviations in O(1) time with no stored samples. The result is numerically more stable than the naive two-pass formula when values span several orders of magnitude.

References

Welford, B. P. (1962). Note on a method for calculating corrected sums of squares and products. Technometrics, 4(3), 419–420.

Implementations

impl RunningStats

pub fn new() -> Self

Creates an empty accumulator.

pub fn push(&mut self, value: f64)

Pushes a new sample and updates all running statistics.

Silently ignores NaN values to avoid poisoning the accumulator.

pub fn count(&self) -> u64

Returns the number of samples pushed.

pub fn is_empty(&self) -> bool

Returns true if no samples have been pushed.

pub fn mean(&self) -> f64

Returns the current running mean, or 0.0 if empty.

pub fn variance(&self) -> f64

Returns the sample variance (divides by n − 1), or 0.0 for < 2 samples.

pub fn population_variance(&self) -> f64

Returns the population variance (divides by n), or 0.0 if empty.

pub fn stddev(&self) -> f64

Returns the sample standard deviation, or 0.0 for < 2 samples.

pub fn population_stddev(&self) -> f64

Returns the population standard deviation, or 0.0 if empty.

pub fn min(&self) -> f64

Returns the minimum value seen, or f64::INFINITY if empty.

pub fn max(&self) -> f64

Returns the maximum value seen, or f64::NEG_INFINITY if empty.

pub fn range(&self) -> f64

Returns the range (max − min), or 0.0 if fewer than two samples.

pub fn cv(&self) -> f64

Returns the coefficient of variation (stddev / |mean|) as a fraction.

Returns f64::NAN when the mean is zero.

pub fn merge(&mut self, other: &Self)

Merges another RunningStats into self using Chan’s parallel algorithm.

After merging, self contains the combined statistics as though all samples from both accumulators had been pushed into a single one.

References

Chan, T. F., Golub, G. H., & LeVeque, R. J. (1979). Updating Formulae and a Pairwise Algorithm for Computing Sample Variances.

pub fn reset(&mut self)

Resets the accumulator to its initial empty state.

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.

impl Default for RunningStats

fn default() -> RunningStats

Returns the “default value” for a type.