Class KMedoidsImputer

java.lang.Object

smile.feature.imputation.KMedoidsImputer

All Implemented Interfaces:

Serializable, Function<Tuple,Tuple>, Transform

public class KMedoidsImputer
extends Object
implements Transform

Missing value imputation by K-Medoids clustering. The k-medoids algorithm is an adaptation of the k-means algorithm. Rather than calculate the mean of the items in each cluster, a representative item, or medoid, is chosen for each cluster at each iteration. The missing values of an instance are replaced the corresponding ones of the nearest medoid.

See Also:

• Serialized Form

Constructor Summary

Constructors

Constructor	Description
KMedoidsImputer(CentroidClustering<Tuple,Tuple> kmedoids)	Constructor.

Method Summary

Modifier and Type	Method	Description
Tuple	apply(Tuple x)
static KMedoidsImputer	fit(DataFrame data, Distance<Tuple> distance, int k)	Fits the missing value imputation values.

Methods inherited from class Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Methods inherited from interface Function

andThen, compose

Methods inherited from interface Transform

andThen, apply, compose

Constructor Details

KMedoidsImputer

public KMedoidsImputer(CentroidClustering<Tuple,Tuple> kmedoids)

Constructor.

Parameters:

kmedoids - the K-Medoids clustering.

Method Details

apply

public Tuple apply(Tuple x)

Specified by:

apply in interface Function<Tuple,Tuple>

fit

public static KMedoidsImputer fit(DataFrame data,
 Distance<Tuple> distance,
 int k)

Fits the missing value imputation values.

Parameters:

data - the training data.

distance - the lambda of distance measure.

k - the number of clusters.

Returns:

the imputer.