|
||||||||||
PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES |
See:
Description
Class Summary | |
---|---|
CloneableText | This class represents vertices in textual form during execution of the different steps of the transitive closure job. |
ClosureDriver | This class encapsulates the execution of MapReduce jobs for calculation of the transitive closure of the vertices within a graph. |
ClusterId | This class represents the unique identifier of a cluster. |
NodeXOrId | Provides a container that contains either a node of a graph or a cluster identifier. |
Step1GroupingComparator | The grouping comparator for the first step provides grouping by the real key of keys which consist of a pair of real key and value. |
Step1Mapper | Assigns the initial cluster identifier to each vertex of the graph. |
Step1MapperBase<KEYIN,VALUEIN> | The Mapper for processing the input and creating the necessary key-value pairs for the next steps of transitive closure creation. |
Step1Pair | A pair of text and cluster identifier. |
Step1Partitioner | Provides distributing keys consisting of real key and value to different reducers only depended on the real key. |
Step1Reducer | Applies the current smallest cluster identifier to each vertex and each other cluster identifier that describes the same cluster. |
Step2GroupingComparator | The grouping comparator for the second step provides grouping by the real key of keys which consist of a pair of real key and value. |
Step2Mapper | Prepares key and value for secondary sort. |
Step2Pair | A pair consisting of a cluster identifier and either a node or an other cluster id. |
Step2Partitioner | Provides distributing keys consisting of real key and value to different reducers only depended on the real key. |
Step2Reducer | The reducer for merging clusters together which describe the same cluster. |
Enum Summary | |
---|---|
ClosureCounter | Contains counter variables necessary for running the transitive closure job. |
Contains the classes for creating the transitive closure of graph based data via Hadoop MapReduce using the transitive closure algorithm developed for the BTC 2010 by Dandy Fenz and Matthias Pohl.
|
||||||||||
PREV PACKAGE NEXT PACKAGE | FRAMES NO FRAMES |