Github user derrickburns commented on a diff in the pull request: https://github.com/apache/spark/pull/2419#discussion_r17640379 --- Diff: mllib/src/main/scala/org/apache/spark/mllib/clustering/package.scala --- @@ -0,0 +1,140 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ + +package org.apache.spark.mllib + +import org.apache.spark.mllib.linalg.Vector +import breeze.linalg.{Vector => BV} +import org.apache.spark.rdd.RDD + +package object base { + + val Zero = 0.0 + val One = 1.0 + val Infinity = Double.MaxValue + val Unknown = -1.0 + + private[mllib] trait FP extends Serializable { + val weight: Double + val raw: BV[Double] + } + + private[mllib] class FPoint(val raw: BV[Double], val weight: Double) extends FP { + override def toString: String = weight + "," + (raw.toArray mkString ",") + lazy val inh = (raw :* (1.0 / weight)).toArray + } + + /** + * A mutable point in homogeneous coordinates + */ + private[mllib] class Centroid extends Serializable { + override def toString: String = weight + "," + (raw.toArray mkString ",") + + def isEmpty = weight == Zero + + var raw: BV[Double] = null + + var weight: Double = Zero + + def add(p: Centroid): this.type = add(p.raw, p.weight) + + def add(p: FP): this.type = add(p.raw, p.weight) + + def sub(p: Centroid): this.type = sub(p.raw, p.weight) + + def sub(p: FP): this.type = sub(p.raw, p.weight) + + def sub(r: BV[Double], w: Double): this.type = { + if (r != null) { + if (raw == null) { + raw = r.toVector :*= -1.0 + weight = w * -1 + } else { + raw -= r + weight = weight - w + } + } + this + } + + def add(r: BV[Double], w: Double) : this.type = { + if (r != null) { + if (raw == null) { + raw = r.toVector + weight = w + } else { + raw += r + weight = weight + w + } + } + this + } + } + + private[mllib] trait PointOps[P <: FP, C <: FP] { --- End diff -- This is the distance function trait. For now, only two implementations are provided. A future pull request will provide others. I have implemented the Kullback-Leibler distance function successfully with this interface. I see no reason why any other distance function cannot be implemented efficiently with this interface, since it provides the opportunity to cache data per point and center.
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org