This repository includes methods for editing to graph classes, solving optimization problems with generic and class-specific algorithms, and "lifting" solutions from subgraphs. If you would like to write your own code that directly invokes these methods, please see example.py. The remainder of this README describes how to use a series of scripts provided for running the entire structural rounding pipeline as might be done in an experimental algorithms setting.
- Python Simply run
make pythonto compile. - C++ Simply run
make cppto compile.
Compile the generator by running make generator.
Then, running ./create_graphs.sh small will create 5 graphs per parameter setting (3600 graphs in total) each with 4 million edges in expectation.
To create smaller graphs, use ./create_graphs.sh test which creates graphs with 100 thousand edges in expectation.
To create larger graphs, use ./create_graphs.sh medium (40 million edges) or ./create_graphs.sh large (200 million edges).
Note that the largest graphs can exceed 600MB in size even with a space efficient encoding.
You can also create different sizes of graphs using ./create_graphs.sh custom <edges> <directory>.
- Python Once compiled, run
python main.py <graphs-directory/> - C++ Once compiled, run
./main <graphs-directory/>or./main <graph.s6>
Once you have created synthetic graphs, you can reproduce our experimental results by running main.py on graphs/small/, graphs/medium/, and graphs/large/.
Important: Structural-Rounding is research software, so you should cite us when you use it in scientific publications!
Please see the CITATION file for citation information.
Structural-Rounding is released under the BSD license; see the LICENSE file. Distribution, modification and redistribution, and incorporation into other software is allowed.