Add DIMACS Clique Challenge Graphs (and their complements)

[math1um]

These are benchmark graphs, often huge, and in an unknown format. But the have been important and are clearly relevant to our project. Put these is a separate list - not the graph_objects list (at least until we get their invariants precomputed and stored).

Here is a description: http://dimacs.rutgers.edu/Challenges/

Here is a description of the graph representation format: http://lcs.ios.ac.cn/~caisw/Resource/about_DIMACS_graph_format.txt

Here are the graphs (a variety of sites are listed, in case some disappear):

1. http://iridia.ulb.ac.be/~fmascia/maximum_clique/
2. http://www.info.univ-angers.fr/pub/porumbel/graphs/
3. http://www.nlsde.buaa.edu.cn/~kexu/benchmarks/graph-benchmarks.htm
4. https://turing.cs.hbg.psu.edu/txn131/clique.html#DIMACS_cliques

[thenealon]

There is a python script for reading these graphs at http://www.dcc.fc.up.pt/~jpp/partition/graphtools.py ; getting from here to sage will still require a little work.

[rbarden]

We need to confirm we have all of the DIMACS graphs from all of the challenges. Just to be thorough.

[math1um]

These should be included with comments about what is known, for instance, size, order AND (especially) the clique number in cases where this is known.

Then users can more easily check their own code against what is known. See for example the data at: http://iridia.ulb.ac.be/~fmascia/maximum_clique/

Is it possible to add doctests to graphs???

[yirkajk]

If you google "docstring", that can help figure out what they can and can't do.

I think the answer is no. We can add docstrings to anything that has a body, like functions and classes. But, we declare graphs as variables, so there's no clear scope in which to leave a docstring.