Struct quantette::KmeansOptions

pub struct KmeansOptions<Color> { /* private fields */ }

A builder struct to specify the parameters for k-means.

Examples

let options = KmeansOptions::new()
    .sampling_factor(0.25)
    .seed(42);

Implementations

impl<Color> KmeansOptions<Color>

pub const fn new() -> Self

Creates a new KmeansOptions with default values.

pub fn sampling_factor(self, sampling_factor: f32) -> Self

Sets the sampling factor which controls what percentage of the image/unique colors to sample.

The default sampling factor is 0.5, that is, to sample half of the input.

Negative or NAN sampling factors will cause k-means to not be run.

pub fn initial_centroids(self, centroids: Centroids<Color>) -> Self

Sets the initial colors/centroids for the k-means algorithm.

By default, these are computed through Wu’s quantiztion algorithm (see the wu module).

pub fn seed(self, seed: u64) -> Self

Sets the seed value for the random number generator.

The default seed is 0.

pub fn batch_size(self, batch_size: u32) -> Self

Sets the batch size for the parallel implementation of the k-means quantizer (minibatch k-means).

Larger batch sizes are faster with dimishing returns. Smaller batch sizes are more accurate but slower to run. The batch is divided evenly among the number of threads.

The default batch size is 4096.

Trait Implementations

impl<Color: Clone> Clone for KmeansOptions<Color>

fn clone(&self) -> KmeansOptions<Color>

Returns a copy of the value.

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

Performs copy-assignment from source.

impl<Color: Debug> Debug for KmeansOptions<Color>

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

Formats the value using the given formatter.

impl<Color> Default for KmeansOptions<Color>

fn default() -> Self

Returns the “default value” for a type.

impl<Color> From<KmeansOptions<Color>> for QuantizeMethod<Color>

fn from(options: KmeansOptions<Color>) -> Self

Converts to this type from the input type.