Struct criterion_stats::Distribution

pub struct Distribution<A>(_);

The bootstrap distribution of some parameter

Methods

impl<A> Distribution<A> where
    A: Float, 

fn confidence_interval(&self, confidence_level: A) -> (A, A) where
    usize: From<A, Output = Result<usize, Error>>, 

Computes the confidence interval of the population parameter using percentiles

Panics

Panics if the confidence_level is not in the (0, 1) range.

fn p_value(&self, t: A, tails: &Tails) -> A

Computes the "likelihood" of seeing the value t or "more extreme" values in the distribution.

Methods from Deref<Target = Sample<A>>

fn as_slice(&self) -> &[A]

Returns a slice that contains all the data points

NOTE: This will be removed in favor of a Deref<Target=[A]> implementation once rust-lang/rust#22257 is fixed.

fn max(&self) -> A

Returns the biggest element in the sample

• Time: O(length)

fn mean(&self) -> A

Returns the arithmetic average of the sample

• Time: O(length)

fn median_abs_dev(&self, median: Option<A>) -> A where
    usize: From<A, Output = Result<usize, Error>>, 

Returns the median absolute deviation

The median can be optionally passed along to speed up (2X) the computation

• Time: O(length)
• Memory: O(length)

fn median_abs_dev_pct(&self) -> A where
    usize: From<A, Output = Result<usize, Error>>, 

Returns the median absolute deviation as a percentage of the median

• Time: O(length)
• Memory: O(length)

fn min(&self) -> A

Returns the smallest element in the sample

• Time: O(length)

fn percentiles(&self) -> Percentiles<A> where
    usize: From<A, Output = Result<usize, Error>>, 

Returns a "view" into the percentiles of the sample

This "view" makes consecutive computations of percentiles much faster (O(1))

• Time: O(N log N) where N = length
• Memory: O(length)

fn std_dev(&self, mean: Option<A>) -> A

Returns the standard deviation of the sample

The mean can be optionally passed along to speed up (2X) the computation

• Time: O(length)

fn std_dev_pct(&self) -> A

Returns the standard deviation as a percentage of the mean

• Time: O(length)

fn sum(&self) -> A

Returns the sum of all the elements of the sample

• Time: O(length)

fn t(&self, other: &Sample<A>) -> A

Returns the t score between these two samples

• Time: O(length)

fn var(&self, mean: Option<A>) -> A

Returns the variance of the sample

The mean can be optionally passed along to speed up (2X) the computation

• Time: O(length)

fn bootstrap<T, S>(&self, nresamples: usize, statistic: S) -> T::Distributions where
    S: Fn(&Sample<A>) -> T,
    S: Sync,
    T: Tuple,
    T: Send,
    T::Distributions: Send,
    T::Builder: Send, 

Returns the bootstrap distributions of the parameters estimated by the 1-sample statistic

• Multi-threaded
• Time: O(nresamples)
• Memory: O(nresamples)

Trait Implementations

impl<A> Deref for Distribution<A>

type Target = Sample<A>

The resulting type after dereferencing.

fn deref(&self) -> &Sample<A>

Dereferences the value.