Struct SampleResult 

pub struct SampleResult<'a> { /* private fields */ }

The results of regressing a single sample.

Implementations

impl<'a> SampleResult<'a>

pub fn from_adhoc(sample_scores: ArrayView1<'a, f64>) -> Self

Create a SampleResult from raw scores, useful for code-reuse during training.

pub fn into_scores(self) -> ArrayView1<'a, f64>

Convert the SampleResult into its underlying raw scores.

pub fn sorted_classes(&self) -> Array1<usize>

An array where every classification index sorted by most to least likely. If you are only interested in the winning classification, simply take the first entry.

Resolving the actual scores or human-readable labels of these classification indices is left to the caller (usually Model).

pub fn probabilities(&self) -> Array1<f64>

An array of probabilities between 0..1 where each value corresponds with the relative probabilities of each classification rather than the absolute score returned by the result matrix. Note that the “probabilities” returned do not correspond to a rigorous mathematical metric, but simply match an intuitive notion of probability. This is generally only useful for reporting to humans.