Struct TopComponentsOptions 

pub struct TopComponentsOptions {
    pub exact_cap: usize,
    pub oversamples: usize,
    pub randomized_iterations: Option<usize>,
    pub tolerance: f64,
    pub max_iterations: usize,
    pub seed: u64,
    pub batch_elements: usize,
    pub dual_max_rows: usize,
}

Settings of TopComponents::fit.

Fields

exact_cap: usize

Above this many features the covariance is not formed (scikit-learn’s randomized solver).

oversamples: usize

Extra columns of the iterated subspace (scikit-learn’s n_oversamples).

randomized_iterations: Option<usize>

Iterations above exact_cap; None: scikit-learn’s (module docs).

tolerance: f64

Relative change of the leading values that ends the exact iteration.

max_iterations: usize

seed: u64

batch_elements: usize

Device elements per batch of rows.

dual_max_rows: usize

At most this many rows (fewer than the features, and rows · features ≤ batch_elements) are decomposed exactly through their rows × rows Gram matrix (module docs); 0: never.

Trait Implementations

impl Clone for TopComponentsOptions

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for TopComponentsOptions

impl Debug for TopComponentsOptions

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

Formats the value using the given formatter.

impl Default for TopComponentsOptions

fn default() -> Self

Returns the “default value” for a type.

impl PartialEq for TopComponentsOptions

fn eq(&self, other: &Self) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for TopComponentsOptions