Struct Comparison 

pub struct Comparison { /* private fields */ }

The scores from compare: every candidate on every metric and split.

Implementations

impl Comparison

pub fn models(&self) -> &[String]

Candidate names, in the order given.

pub fn metrics(&self) -> &[String]

Metric names, in the order given.

pub fn n_splits(&self) -> usize

Number of splits.

pub fn split_scores(&self, model: usize, metric: usize) -> &[f64]

Scores of model on metric, one per split (indices into models and metrics).

pub fn mean(&self, model: usize, metric: usize) -> f64

Mean score over the splits.

pub fn std_error(&self, model: usize, metric: usize) -> f64

Standard error of mean: the splits’ standard deviation over √k (NaN with one split), as cv.glmnet and tidymodels report it.

pub fn best(&self, metric: usize) -> usize

The candidate with the lowest mean on metric.

pub fn difference_std_error(&self, model: usize, metric: usize) -> f64

Standard error of each candidate’s per-split difference from the best on metric (0 for the best itself). Splits are shared, so the paired difference is far less noisy than either mean: a candidate within about two of these of the best is not clearly worse.

Trait Implementations

impl Clone for Comparison

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Comparison

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

Formats the value using the given formatter.

impl PartialEq for Comparison

fn eq(&self, other: &Self) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for Comparison