QTreeVine

Sparse linear algebra in Vine

How to run

Explore options of run in documentation.

1. Use official instructions to install vine.
2. vine run src/tests.vi --lib src/lib to run all tests.
3. vine run src/run_BFS.vi --lib src/lib to run BFS example.
4. vine run src/run_SSSP.vi --lib src/lib to run SSSP example.
5. vine run src/run_TriangleCount.vi --lib src/lib to run TriangleCount example.

For semi-automatic run you can use run.py:

python3 run.py {BFS,SSSP,TriangleCount} [matrix_file]
               [-r RUNS] [-w WORKERS_START WORKERS_END] [-H HEAP_GB]

• {BFS,SSSP,TriangleCount} — algorithm to run (required)
• matrix_file — path to .mtx file (optional; overrides value in source)
• -r RUNS — number of runs per configuration (default: 1)
• -w WORKERS_START WORKERS_END — worker range (end is exclusive) (default: 4 5)
• -H HEAP_GB — heap size in GB, converted to bytes and passed to vine (optional; default: vine's built-in limit)

Logs are placed in logs_{algo}[_{matrix_stem}]/ with files named {workers}[_H{heap}G].log. Run output is appended to the respective log file. The source file is temporarily modified to set matrix_file if a matrix override is given, and restored after all runs.