Cluster Kmeans Params Module

Rust module: cuvs::cluster::kmeans::params

Source: rust/cuvs/src/cluster/kmeans/params.rs

Params

1
pub struct Params(pub ffi::cuvsKMeansParams_t);

Methods

new

1
pub fn new() -> Result<Params> { ... }

Returns a new Params

Source: rust/cuvs/src/cluster/kmeans/params.rs:15

set_metric

1
pub fn set_metric(self, metric: DistanceType) -> Params { ... }

DistanceType to use for fitting kmeans

Source: rust/cuvs/src/cluster/kmeans/params.rs:24

set_n_clusters

1
pub fn set_n_clusters(self, n_clusters: i32) -> Params { ... }

The number of clusters to form as well as the number of centroids to generate (default:8).

Source: rust/cuvs/src/cluster/kmeans/params.rs:32

set_max_iter

1
pub fn set_max_iter(self, max_iter: i32) -> Params { ... }

Maximum number of iterations of the k-means algorithm for a single run.

Source: rust/cuvs/src/cluster/kmeans/params.rs:40

set_tol

1
pub fn set_tol(self, tol: f64) -> Params { ... }

Relative tolerance with regards to inertia to declare convergence.

Source: rust/cuvs/src/cluster/kmeans/params.rs:48

set_n_init

1
pub fn set_n_init(self, n_init: i32) -> Params { ... }

Number of instance k-means algorithm will be run with different seeds.

Source: rust/cuvs/src/cluster/kmeans/params.rs:56

set_oversampling_factor

1
pub fn set_oversampling_factor(self, oversampling_factor: f64) -> Params { ... }

Oversampling factor for use in the k-means|| algorithm

Source: rust/cuvs/src/cluster/kmeans/params.rs:64

set_batch_samples

1
pub fn set_batch_samples(self, batch_samples: i32) -> Params { ... }

Source: rust/cuvs/src/cluster/kmeans/params.rs:77

set_batch_centroids

1
pub fn set_batch_centroids(self, batch_centroids: i32) -> Params { ... }

if 0 then batch_centroids = n_clusters

Source: rust/cuvs/src/cluster/kmeans/params.rs:84

set_hierarchical

1
pub fn set_hierarchical(self, hierarchical: bool) -> Params { ... }

Whether to use hierarchical (balanced) kmeans or not

Source: rust/cuvs/src/cluster/kmeans/params.rs:92

set_hierarchical_n_iters

1
pub fn set_hierarchical_n_iters(self, hierarchical_n_iters: i32) -> Params { ... }

For hierarchical k-means , defines the number of training iterations

Source: rust/cuvs/src/cluster/kmeans/params.rs:100

Source: rust/cuvs/src/cluster/kmeans/params.rs:11