Trait Classifier 

pub trait Classifier<Features, Label>
where
    Label: Clone,
{
    // Required methods
    fn labels(&self) -> &[Label];
    fn predict_proba<I>(&self, arr: I) -> Option<Array2<f64>>
       where I: Iterator<Item = Features>;

    // Provided method
    fn predict<I>(&self, arr: I) -> Option<Vec<Label>>
       where I: Iterator<Item = Features> { ... }
}

Trait to interface with a fitted classification model

Required Methods

fn labels(&self) -> &[Label]

Labels on which the model is fitted.

fn predict_proba<I>(&self, arr: I) -> Option<Array2<f64>>

where I: Iterator<Item = Features>,

Estimates likelihood of each class per record. Rows correspond to each record, columns are in the same order as label function.

Provided Methods

fn predict<I>(&self, arr: I) -> Option<Vec<Label>>

where I: Iterator<Item = Features>,

Provided function which returns the most likely class per record based on the results of predict_proba().

Dyn Compatibility

This trait is not dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety", so this trait is not object safe.

Implementors

impl<Input: Estimatable, Label: Clone> Classifier<Input, Label> for GaussianNB<Input, Label>