Struct linfa_clustering::KMeansHyperParamsBuilder

pub struct KMeansHyperParamsBuilder<F: Float, R: Rng> { /* fields omitted */ }

An helper struct used to construct a set of valid hyperparameters for the K-means algorithm (using the builder pattern).

Implementations

impl<'a, F: Float, R: Rng + SeedableRng + Clone> KMeansHyperParamsBuilder<F, R>

pub fn fit<D: Data<Elem = F>, T>(
    self,
    dataset: &DatasetBase<ArrayBase<D, Ix2>, T>
) -> Result<KMeans<F>, KMeansError>

Shortcut for .build().fit()

impl<F: Float, R: Rng + Clone> KMeansHyperParamsBuilder<F, R>

pub fn n_runs(self, n_runs: usize) -> Self

Set the value of n_runs.

The final results will be the best output of n_runs consecutive runs in terms of inertia (sum of squared distances to the closest centroid for all observations in the training set). For incremental K-means, only the initialization algorithm will be run multiple times to pick the best starting centroids.

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

Set the value of max_n_iterations.

We exit the training loop when the number of training iterations exceeds max_n_iterations even if the tolerance convergence condition has not been met. Not used when training incrementally.

pub fn tolerance(self, tolerance: F) -> Self

Set the value of tolerance.

The training is considered complete if the euclidean distance between the old set of centroids and the new set of centroids after a training iteration is lower or equal than tolerance.

pub fn init_method(self, init: KMeansInit<F>) -> Self

Set the value of init.

The initialization method is the function that determines the initial values of the cluster centroids before the iterative training process. The default value is KMeansPlusPlus.

pub fn build(self) -> KMeansHyperParams<F, R>

Return an instance of KMeansHyperParams after having performed validation checks on all the specified hyperparameters.

Panics if any of the validation checks fails.