Struct ClassificationMetrics 

pub struct ClassificationMetrics {
    pub total: usize,
    pub correct: usize,
    pub class_tp: HashMap<String, usize>,
    pub class_fp: HashMap<String, usize>,
    pub class_fn: HashMap<String, usize>,
    pub class_support: HashMap<String, usize>,
}

Metrics for text classification tasks.

Supports multi-class classification with macro/micro/weighted averaging.

Fields

total: usize

Total number of examples

correct: usize

Number of correct predictions

class_tp: HashMap<String, usize>

Per-class true positives

class_fp: HashMap<String, usize>

Per-class false positives

class_fn: HashMap<String, usize>

Per-class false negatives

class_support: HashMap<String, usize>

Per-class support (total examples per class)

Implementations

impl ClassificationMetrics

pub fn new() -> Self

Create new empty metrics.

pub fn add(&mut self, predicted: &str, actual: &str)

Add a prediction to the metrics.

pub fn accuracy(&self) -> f64

Overall accuracy.

pub fn macro_precision(&self) -> f64

Macro-averaged precision (unweighted average across classes).

pub fn macro_recall(&self) -> f64

Macro-averaged recall (unweighted average across classes).

pub fn macro_f1(&self) -> f64

Macro-averaged F1 score.

pub fn micro_precision(&self) -> f64

Micro-averaged precision (aggregate TP/FP across classes).

pub fn micro_recall(&self) -> f64

Micro-averaged recall (aggregate TP/FN across classes).

pub fn micro_f1(&self) -> f64

Micro-averaged F1 score.

pub fn weighted_f1(&self) -> f64

Weighted F1 score (weighted by class support).

pub fn class_precision(&self, class: &str) -> f64

Precision for a specific class.

pub fn class_recall(&self, class: &str) -> f64

Recall for a specific class.

pub fn class_f1(&self, class: &str) -> f64

F1 for a specific class.

pub fn classes(&self) -> Vec<&String>

Get all class labels.

Trait Implementations

impl Clone for ClassificationMetrics

fn clone(&self) -> ClassificationMetrics

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ClassificationMetrics

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

Formats the value using the given formatter.

impl Default for ClassificationMetrics

fn default() -> ClassificationMetrics

Returns the “default value” for a type.