com.cloudera.oryx.app.kmeans

Class KMeansUtils

java.lang.Object

com.cloudera.oryx.app.kmeans.KMeansUtils

public final class KMeansUtils
extends Object

K-means-related utility methods for the app tier.

Method Summary

Modifier and Type	Method and Description
static void	checkUniqueIDs(Collection<ClusterInfo> clusters)
static Pair<ClusterInfo,Double>	closestCluster(List<ClusterInfo> clusters, DistanceFn<double[]> distanceFn, double[] vector)
static double[]	featuresFromTokens(String[] data, InputSchema inputSchema)

Methods inherited from class java.lang.Object

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

Method Detail

closestCluster

public static Pair<ClusterInfo,Double> closestCluster(List<ClusterInfo> clusters,
                                                      DistanceFn<double[]> distanceFn,
                                                      double[] vector)

Parameters:

clusters - all clusters

distanceFn - distance function between clusters

vector - point to measure distance to

Returns:

cluster whose center is closest to the given point

featuresFromTokens

public static double[] featuresFromTokens(String[] data,
                                          InputSchema inputSchema)

Parameters:

data - tokenized input

inputSchema - input schema to apply to input

Returns:

input parsed as numeric values

checkUniqueIDs

public static void checkUniqueIDs(Collection<ClusterInfo> clusters)

Parameters:

clusters - clusters to check the IDs of

Throws:

IllegalArgumentException - if any ClusterInfo.getID() is duplicated