Struct Emv

pub struct Emv { /* private fields */ }

Exponential moving variance. An estimate of the (max-min) spread for the last N items. (Work in progress!)

This effectively gives a spread (max-min) over a window size N without the need to store N values.

Welford’s algorithm is used to calculate the running variance,

https://en.wikipedia.org/wiki/Algorithms_for_calculating_variance#Welford's_online_algorithm

The spread_factor called the “the range rule of thumb”, and is used to deduce a max-min spread from the rolling standard deviation.

https://en.wikipedia.org/wiki/68%E2%80%9395%E2%80%9399.7_rule

Implementations

impl Emv

pub fn with_factor(virtual_window_size: usize, spread_factor: f64) -> Self

The smoothing factor alpha is calculated from the virtual_window_size and the spread_factor represents a multiplier to estimate max-min from standard deviation. For Gaussian distributions, a factor of 6 assumes max-min = 6 standard deviations, a factor of 4 assumes spread = max-min = 4 standard deviations.

pub fn new(virtual_window_size: usize) -> Self

The smoothing factor alpha is calculated from the virtual_window_size and the spread factor is set to be 6 (using the rule of thumb for Gaussian distributions). For Gaussian distribution cdf(3) - cdf(-3) = 99.7%, so 3-(-3) is appropriate.

pub fn std_deviation(&self) -> f64

pub fn mean(&self) -> f64

Will return NaN before window size data items have been sampled

Trait Implementations

impl Clone for Emv

fn clone(&self) -> Emv

Returns a copy of the value.

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

Performs copy-assignment from source.

impl Debug for Emv

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

Formats the value using the given formatter.

impl Display for Emv

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

Formats the value using the given formatter.

impl Metric<f64> for Emv

type Output = f64

The type of elements recorded by the metric. For a delta of last vs current, this might be an n-dimentional vector. It typically needs to be an Owned type, rather than a reference.

fn observe_opt(&mut self, x: f64) -> Option<Self::Output>

Similar to observe except None is returned instead of f64::NAN to indicate not enough data collected. Useful if you like

fn observe(&mut self, x: Input) -> f64

where Self::Output: Into<f64>,

Records the value, and return the calculated metric. If there are not enough samples to calculate the metric, f64::NAN is returned, which will always compare false. So a tolerance check observe(x) < 0.0001 will fail until enough samples have been collected. If you don’t like NANs, then Metric::observe_opt is your friend.

impl StatefulMetric<f64> for Emv

fn store(&mut self, x: f64)

Stores the value, and likely discards older values.

fn value_opt(&self) -> Option<f64>

The calculated value.

fn observe_opt(&mut self, x: Input) -> Option<Self::Output>

fn value(&self) -> f64

where Self::Output: Into<f64>,

The calculated value.