Build Instructions

Table of Contents Requirements List of modules Build Profiles Building executable jar with all dependencies Building a jar for spark submission

Requirements

• Newer JDK (1.7+) : Tested on JDK 7
• Newer Maven (3.x+): Tested on Maven 3.3x
• Working Internet Connection

List of modules

This project contains following modules

• autoext - The core module
• autoext-spark - the module for performing distributed clustering on Apache Spark
• apted - A faster Tree Edit Distance (TED) Implementation
• visuals - Web interface for visualizations

Build Profiles

• Executable jar with all dependencies (default)
• spark-submit - profile for packaging jar for submitting to spark vis spark-submit command

Building executable jar with all dependencies

• go to the root of the project cd
• mvn clean compile package

On success, the executable jar shall be at found at autoext-spark/target/autoext-spark-xx.jar

To run this jar : java -jar autoext-spark/target/autoext-spark-xx.jar

Building a jar for spark submission

• mvn clean compile package -Pspark-submit

This build excludes spark and scala dependencies from package. On success, the jar can be found at autoext-spark/target/autoext-spark-xx-SNAPSHOT-submit-{spark.version}_{scala.version}.jar

To run this jar, use spark-submit command

For quick start tutorial visit Clustering tutorial

NOTE:

 If you have trouble compiling the project, make sure your version of JDK, Scala and Spark are compatible.