Function parsnip::fbeta_score

pub fn fbeta_score(
    pred: &[u64], 
    actual: &[u64], 
    beta: f32, 
    average: Option<String>
) -> f32

The recall of a dataset Returns the hamming loss which is the percentage of items which are misclassified [0, 1]

Supports macro and weighted averages

use parsnip::fbeta_score;
 
let actual = vec![0, 1, 2, 0, 1, 2];
let pred = vec![0, 2, 1, 0, 0, 1];
 
assert_eq!(fbeta_score(&pred, &actual, 0.5, Some("macro".to_string())), 0.23809524);
assert_eq!(fbeta_score(&pred, &actual, 0.5, Some("weighted".to_string())), 0.23809527);