##installation
Graph analysis using the Giraph apache incubator project
To compile:
mvn package
Running the analysis requires a compiled trunk version of Giraph:
git clone git@github.com:apache/giraph.git
cd giraph
mvn install
##Analysis ###pagerank (using an edgelist of strings as input)
hadoop jar target/Analysis-0.0.1-SNAPSHOT-jar-with-dependencies.jar\
org.apache.giraph.GiraphRunner org.data2semantics.giraph.pagerank.PageRankComputation \
-eif org.data2semantics.giraph.io.EdgeListStringReader \
-of org.apache.giraph.io.formats.IdWithValueTextOutputFormat \
-mc org.data2semantics.giraph.pagerank.RandomWalkVertexMasterCompute \
-wc org.data2semantics.giraph.pagerank.RandomWalkWorkerContext
-op <outputPath> \
-eip <input edge list> \
-w <number of workers>
This executed pagerank for inputfile <input edge list>, writes output to directory <outputPath>, and uses this amount of workers: <number of workers>
###pagerank (using an edgelist of longs as input)
hadoop jar target/Analysis-0.0.1-SNAPSHOT-jar-with-dependencies.jar\
org.apache.giraph.GiraphRunner org.data2semantics.giraph.pagerank.numerical.PageRankComputation\
-eif org.data2semantics.giraph.io.EdgeListLongReader\
-of org.apache.giraph.io.formats.IdWithValueTextOutputFormat
-mc org.data2semantics.giraph.pagerank.numerical.RandomWalkVertexMasterCompute\
-wc org.data2semantics.giraph.pagerank.numerical.RandomWalkWorkerContext\
-op <outputPath> \
-eip <input edge list> \
-w <number of workers>
This executed pagerank for inputfile <input edge list>, writes output to directory <outputPath>, and uses this amount of workers: <number of workers>
###outdegree
hadoop jar target/Analysis-0.0.1-SNAPSHOT-jar-with-dependencies.jar\
org.apache.giraph.GiraphRunner org.data2semantics.giraph.pagerank.SimpleOutDegreeCountComputation \
-eif org.data2semantics.giraph.io.EdgeListReader \
-of org.apache.giraph.io.formats.IdWithValueTextOutputFormat \
-op <outputPath> \
-eip <input edge list> \
-w <number of workers>
This executed pagerank for inputfile <input edge list>, writes output to directory <outputPath>, and uses this amount of workers: <number of workers>
###outdegree
hadoop jar target/Analysis-0.0.1-SNAPSHOT-jar-with-dependencies.jar\
org.apache.giraph.GiraphRunner org.data2semantics.giraph.pagerank.SimpleInDegreeCountComputation \
-eif org.data2semantics.giraph.io.EdgeListReader \
-of org.apache.giraph.io.formats.IdWithValueTextOutputFormat \
-op <outputPath> \
-eip <input edge list> \
-w <number of workers>
This executed pagerank for inputfile <input edge list>, writes output to directory <outputPath>, and uses this amount of workers: <number of workers>