This version contains many bug fixes, and switches the underlying graph representation from a boost's adjacency_list to a much simpler, vector-based custom class which uses only about half as much memory. Get it in the download section.

This release contains mainly bug fixes, as well as some improvements to blockmodel inference, and the collection module, as well as a new cookbook section in the documentation about animations.

This release contains several improvements and bug fixes. Highlights include support for blockmodel inference, a collection of useful networks in the new collection module, pie chart vertex shape in graph_draw, several python3 fixes, and many other bug fixes.

This release contains improvements and bug fixes. Highlights include new Katz and HITS centrality algorithms (in centrality), and make_maximal_planar (in topology).

This release contains major improvements and bug fixes. Highlights include a completely new graph_draw() function, which is based on  cairo and GTK+3, has an interactive window, and no longer requires graphviz (the old function is still available as graphviz_draw()), implementation of the SFDP mulilevel layout algorithm, and support for reading and writing in  GML format. Get it in the download section.

This release contains several enhancements and bug fixes, such as a comprehensive improvement of the random_graph() and random_rewire() functions (with the inclusion of blockmodel generation), the implementation of eigenvector centrality and the similarity() function. Array access to the property maps of filtered graphs has also been improved with the addition of the fa and ma attributes. Get it in the download section.

This release includes some bug fixes, improvements to the PageRank code, and an implementation of BGL's Fruchterman-Reingold layout algorithm. Get it in the download section.

This release is a just a follow-up to 2.2.12, with some compilation bugfixes which affect some versions of Boost. Get it in the download section.

This version includes bug fixes, a new label_largest_component() function, and better support for MacOS X. Get it in the download section.

This version includes bug fixes, and a new feature called "graph views", which provides a very convenient and powerful interface to the graph filtering functionality (see the documentation for more details). Get it in the download section.

This version includes several bugfixes, specially to the flow module. Get it in the download section.

This version includes several bugfixes, and a brand new search module. Get it in the download section.

This version includes several bugfixes, and new graph generators. Get it in the download section.

Important note

The graph-tool website permanently moved to  http://graph-tool.skewed.de, and the documentation now reflects this. Please update your bookmarks!

This version includes several bugfixes, and a slight modification of the binning semantics of the various histogram functions, so that it corresponds to the numpy  histogram function. Get it in the download section.

This version includes some bugfixes, and a significative speedup of the implementation of random graph generation. Get it in the download section.

This version includes several bugfixes, a revamped implementation of random graph generation and rewiring and a new shortest path function. Get it in the download section.

This version includes bugfixes (including compilation fixes for MacOS) and an implementation of subgraph isomorphism. Get it in the download section.

This version includes periodic Delaunay triangulation. Get it in the download section.

This version includes bug-fixes and an additional "erdos" strategy in the random_rewire funtion. Get it in the download section.

This version includes 2D and 3D point triangulation, and planarity test. Get it in the download section.

This is mainly a bug-fix release. Get it in the download section.

This version includes a shortest distance function, and several bug fixes. Get it in the download section.

A new stable version of graph-tool was released. It marks a long transition from a stand-alone program to a full python module. It is pointless to summarize all the differences, since everything changed. The new version is a lot more comfortable to work with, and the internals are a lot cleaner. Everything integrates nicely with scipy/numpy, and I hope graph-tool will be a nice addition to this framework.