From 023702c90da7c421e1e339e14878bccebbb26e24 Mon Sep 17 00:00:00 2001 From: "Joseph E. Gonzalez" Date: Mon, 19 Aug 2013 13:32:54 -0700 Subject: [PATCH] Updating documentation. --- core/src/main/scala/spark/rdd/IndexedRDD.scala | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/core/src/main/scala/spark/rdd/IndexedRDD.scala b/core/src/main/scala/spark/rdd/IndexedRDD.scala index b2a2180b8f..d19d220ec2 100644 --- a/core/src/main/scala/spark/rdd/IndexedRDD.scala +++ b/core/src/main/scala/spark/rdd/IndexedRDD.scala @@ -35,18 +35,20 @@ import spark.Partitioner._ /** - * And index RDD + * The BlockIndex is the internal map structure used inside the index + * of the IndexedRDD. */ class BlockIndex[@specialized K: ClassManifest] extends JHashMap[K,Int] -//type BlockIndex[@specialized K: ClassManifest] = JHashMap[K,Int] /** - * An IndexedRDD is an RDD[(K,V)] where each K is unique. + * An IndexedRDD[K,V] extends the RDD[(K,V)] by pre-indexing the keys and + * organizing the values to enable faster join operations. + * + * In addition to providing the basic RDD[(K,V)] functionality the IndexedRDD + * exposes an index member which can be used to "key" other IndexedRDDs * - * The IndexedRDD contains an index datastructure that can - * be used to accelerate join and aggregation operations. */ class IndexedRDD[K: ClassManifest, V: ClassManifest]( val index: RDD[ BlockIndex[K] ],