Dataset-preparation.txt

Download links for datasets:
LiveJournal - http://snap.stanford.edu/data/soc-LiveJournal1.html
com-orkut - http://snap.stanford.edu/data/com-Orkut.html
twitter -  http://an.kaist.ac.kr/~haewoon/release/twitter_social_graph/twitter_rv.tar.gz
pld_arc - http://data.dws.informatik.uni-mannheim.de/hyperlinkgraph/2012-08/pld-arc.gz
kron - Refer the issue on GAP benchmark for "Generation of kronecker graphs" (Also see the note below)
road - http://www.dis.uniroma1.it/challenge9/data/USA-road-d/USA-road-d.USA.gr.gz
urand - Refer bench.mk inside GAP benchmark original repo by Scott (Also see the note below)
soc-pokec - http://snap.stanford.edu/data/soc-pokec-relationshps.txt.gz

Extract the .zip or .gz files

Once the .txt file is downloaded convert them to .el files
mv <file-name>.txt <file-name>.el


Convert the .el file to .sg file using converter.cc of GAP benchmark:
./converter -f <file-name>.el -b <file-name>.sg

Convert the .el file to .wsg file using converter.cc of GAP benchmark:
./converter -f <file-name>.el -wb <file-name>.wsg

Note:
These values are calculated as per DROPLET HPCA'19 paper by Abanti Basak
1. Generation of kron dataset commands:
    .sg file:
        ./converter -g24 -k16 -b <file-name>.sg
    .wsg file:
        ./converter -g24 -k16 -wb <file-name>.wsg
      
2. Generation of urand dataset commands:
    .sg file:
        ./converter -u23 -k16 -b <file-name>.sg
    .wsg file:
        ./converter -u23 -k16 -wb <file-name>.wsg