How to use the benchmark suite:

1) Overview about available benchmarks
#>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu 
  org.semanticweb.bench.Menu <benchmark>
  where <benchmark> one of
        StmtInsert   Benchmark the time to insert X stmt
        Lookups      Benchmark the lookup time and quality for BGPs
        WebFetch     TBD
        GenerateBGP  Randomly select BGPs from data

2)GenerateBGPs
#>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu GenerateBGPs 
  usage: parameters:
  -h          print help
  -i <arg>    input location
  -if <arg>   input format (idx, dir)
  -is <arg>   number of estimated input stmts (estimation)
  -o <arg>    output file
  -p <arg>    Pattern for BGP (-1,0). -1 = variable, 0 = select, sperated
             by ",", e.g. only predicates (-1,0,-1)
  -s <arg>    number of patterns

 Example
 # Generate a sampling of 100 items for the BGP class ?var <p> <o>
 # The sampling should be taken from files in the folder "input/linked-data" which contains around 110000 stms (this 
 # is important to get a equal distributed sampling)
 #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu GenerateBGPs -i input/linked-data/ -if dir -is 110000 -s 100 -p -1,0,0 -o input/100_po_bgps

3)StmtInsert
 #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu StmtInsert
 usage: parameters:
 -b <arg>    number of buckets [default 1000]
 -f <arg>    fanout value [default 20]
 -h          print help
 -i <arg>    input location
 -if <arg>   input format (idx, dir)
 -o <arg>    output location
 -s          serialise the create qtree (flag)
 
 The output folder will contain time and used-memory measures and the relevant gnuplot files to generate some graphs
 Example
 #Insert all stmts from files in input/linked-data into a QTree with bucket size 10000 and fanout of 30, add the end serialise the QTree into the output folder
 #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu StmtInsert -i input/linked-data/ -o tmp -f 30 -b 10000 -if dir -s 

4)Lookups
 #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu LookupsMissing required argument i/o/if!
 -d <arg>    original data location
 -h          print help
 -i <arg>    qtree location
 -if <arg>   input format of the original data(idx, dir)
 -l <arg>    file containing new line seperated basic graph patterns
 -o <arg>    output location

 Output folder contains the time and quality measure for each BGP lookup and the average for all lookups
 Example
 #Test the 100 random selected BGPs for the created QTree. The data to verify how many sources were really required are taken from input/linked-data 
 #>java -cp dist/semqtree.jar org.semanticweb.bench.menu.Menu Lookups -i tmp/_qtrees/qtree4all--hprefix_b10000_f30.ser -if dir -d input/linked-data/ -l input/100_spo_bgps -o tmp/