This fixes the expected values for several doctests, after
random_graph() re-factoring.

Small fixes of some examples in the docstrings.

This completes all the documentation up to this point.

In both functions we can save some time by hashing the subgraphs found,
according to their degree distribution signature (instead of number of
edges). This avoids a large number of useless exactness or isomorphism
comparisons.

This also removes the "seed" parameter, which is redundant to
the numpy.random.seed() function.

Mostly in the flow and spectral modules.

Load 'topology' instead of 'misc'.

The RNG object must be passed by _reference_ not by value, as is the
default! Otherwise the RNG state does not change...

This modified several docstrings and introduces a quick-start guide.

The function directed() no longer exists, is_directed() should be used
instead.

This includes the function motif_significance() into the clustering
module.

This implements an efficient motif counting algorithm.

Reduce duplication of code for loading C++ modules.

Also, get rid of g.underlying_graph(), and rely on "private" variable
g._Graph__graph instead.

A new 'clustering' sub-module was created.

The whole histogram code has been redone, and the code has been
simplified. The three-point vertex-edge-vertex correlation has been
scrapped, since it's not frequently used, and would make compilation
even more expensive.

This also adds some missing files to the generation routine.