Analyze

Compile

    sbt assembly

The result jar will be located under ./target/scala-2.10/analyze-assembly-1.0.jar

Execution

    analyze (who.cares.the.version - number RC)
    Usage: analyze [options]

      -a <value> | --alpha <value>
            Damping factor of PR & PPR default 0.15
      -i <value> | --iter <value>
            Max iteration number of PR & PPR default 15
      -k <value> | --topk <value>
            TopK elements that will return default 20
      -c <value> | --c <value>
            Parameter for backward PPR, part of decay factor \beta default 50
      --lpaStart <value>
            Min iteration of LPA
      --lpaStep <value>
            Step of LPA iteration
      --lpaEnd <value>
            Max iteration of LPA community detection
      -q <value> | --query <value>
            Query id, usually the starting point of PPR, default is 0l
      -e <value> | --edge <value>
            Path of edge file, should be on hdfs
      -t <value> | --title <value>
            Path of title-id map, should be on local disk
      -o <value> | --output <value>
            Output path, should be a local location
      -n <value> | --ncore <value>
            Number of cores for Spark

Example

    spark-submit --class edu.nd.dsg.bshi.Main --master yarn-client --num-executors 6 \ 
     --driver-memory 12g --executor-memory 14g --executor-cores 8 analyze-assembly-1.0.jar \
     -e hdfs://dsg1.crc.nd.edu/user/bshi/dblp/citation_edge.txt -o ./text_result.csv --lpaStart 10 --lpaEnd 11