Release date: 11 April 2019
Supports Python 3.5, 3.6 and 3.7. This is our first Python 3 only release.
We’re happy to announce the release of NetworkX 2.3! NetworkX is a Python package for the creation, manipulation, and study of the structure, dynamics, and functions of complex networks.
This release is the result of 6 months of work with over 92 pull requests by 30 contributors. Highlights include:
Dropped support for Python 2. We are no longer supporting Python 2.7 and we will start changing code to take advantage of Python 3 features we couldn’t before.
Added some Moral Graph analysis functions.
Enable matplotlib drawing using curved arrows via connectionstyle parameter.
Remove ticks and axes labels from matplotlib plots.
Two new generators of Harary Graphs.
Added Dual Barabasi-Albert model
Added VoteRank algorithm
Added Equitable coloring algorithms
Added planar layout algorithms
Les Miserables network example
Change default colors to be color-blind friendly
Many bug fixes and documentation improvements
Speed up of simple_cycles
Improvements for reading various formats like GML, GEXF, Graphml
Allow subclassing to access node_attr_dict_factory
The G.fresh_copy() mechanism for creating an empty_graph of the same type (introduced in v2.0) does not playing nicely with pickle and others. So, we have removed the code that caused a need for that. Instead you should use the more natural G.__class__() syntax to get an empty_graph of the same type as G.
The Graph.fresh_copy() method should now use Graph.__class__()
ReverseView class removed in favor of reverse_view() function.
Clayton A Davis
Øyvind Heddeland Instefjord
Warren W. Kretzschmar
Edward L Platt