Struct Leaderboard 

pub struct Leaderboard {
    pub results: Vec<EvalResult>,
    pub primary_metric: Metric,
}

Leaderboard for comparing multiple models

Fields

results: Vec<EvalResult>

Evaluation results for each model

primary_metric: Metric

Primary metric for ranking

Implementations

impl Leaderboard

pub fn new(primary_metric: Metric) -> Self

Create a new leaderboard

pub fn add(&mut self, result: EvalResult)

Add evaluation result

pub fn sort(&mut self)

Sort by primary metric

N-06 (Meyer DbC): Models with missing scores sort last, not to an arbitrary position. NEG_INFINITY for “higher is better” and INFINITY for “lower is better” ensure worst-possible semantics.

pub fn sort_by(&mut self, metric: Metric)

Sort by a specific metric

pub fn best(&self) -> Option<&EvalResult>

Get best model by primary metric

pub fn print(&self)

Print formatted leaderboard to stdout (Mieruka - visual control)

pub fn to_markdown(&self) -> String

Export as markdown table

Trait Implementations

impl Clone for Leaderboard

fn clone(&self) -> Leaderboard

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for Leaderboard

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

Formats the value using the given formatter.

impl Display for Leaderboard

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

Formats the value using the given formatter.