Struct TfidfTransformer 

pub struct TfidfTransformer {
    pub idf: Vec<f64>,
    pub smooth_idf: bool,
    /* private fields */
}

Transforms a count matrix into a TF-IDF weighted matrix.

IDF formula (with smooth_idf = true): idf(t) = ln((1 + n) / (1 + df(t))) + 1

Each row is L2-normalised after weighting.

Fields

idf: Vec<f64>

Per-term IDF values.

smooth_idf: bool

Whether to smooth IDF by adding 1 to numerator and denominator.

Implementations

impl TfidfTransformer

pub fn new(smooth_idf: bool) -> Self

Create a new transformer. smooth_idf = true avoids division by zero for terms seen in all documents.

pub fn fit(&mut self, count_matrix: &[Vec<f64>]) -> Result<()>

Compute IDF values from a count matrix (rows = documents, cols = terms).

pub fn transform(&self, count_matrix: &[Vec<f64>]) -> Result<Vec<Vec<f64>>>

Apply TF-IDF weighting and L2 normalisation.

pub fn fit_transform( &mut self, count_matrix: &[Vec<f64>], ) -> Result<Vec<Vec<f64>>>

Fit then transform in one step.

Trait Implementations

impl Clone for TfidfTransformer

fn clone(&self) -> TfidfTransformer

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TfidfTransformer

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

Formats the value using the given formatter.