Posted on by admin

Baobab

ogdf.v2015.05Download

Today, we released the new version v2015.05 (Baobab) of OGDF! The release introduces various new algorithms and modules but also improves usability and elegance. To accomplish the latter, some changes had to be made that break compatibility to Sakura. Also, COIN-OR and ABACUS are now included in the OGDF package to make every feature work out-of-the-box.

In parallel to the release of Baobab, we uploaded a new snapshot of OGDF. It marks the start in using C++11 features and the CMake build system. Note that you will need much more recent compilers to compile the snapshot than you would need for the Baobab release or former releases.

Highlights

  • COIN-OR and ABACUS are integrated in the package for easier installation
  • improvements to GraphAttributes class
    • restructuring
    • more consistent naming scheme
    • improved color management (Color class instead of strings)
    • basic 3d support
  • new LayoutStandards: changeable default parameters (sizes, distances, colors) that allow more comparable drawings
  • basic support for hypergraphs (Hypergraph class)
  • reader and writer for fileformats
    • new support for DL, DOT, GDF, GEXF, GraphML, Tulip graph formats
    • new support for Bench and PLA hypergraph formats
    • improved GML and OGML parser
    • reading and writing files is now handled by GraphIO class
  • new layered crossing minimization for Sugiyama algorithm: grid sifting and global sifting heuristics (GridSifting, GlobalSifting; Bachmaier et al, 2011)
  • new EmbedderOptimalFlexDraw class for bend minimization in planar, orthogonal drawings (Bläsius et al, 2012)
  • new force-directed layout BertaultLayout that preserves edge-crossing properties (Bertault, 2000)
  • new parallelized version of FastPlanarSubgraph
  • new methods for some (restricted) upward planarity problems (UpwardPlanarity class)
  • new classes for stress majorization (StressMinimization), pivot multi-dimensional scaling (PivotMDS)
  • revised code for edge insertion into planar graphs (EdgeInsertionModule, FixedEmbeddingInserter, VariableEmbeddingInserter*, SubgraphPlanarizer classes)
  • new Graph::CCsInfo class to store information about connected components
  • new LCA class to compute lowest common ancestors in arborescences with linear preprocessing time and constant query time (Bender and Farach-Colton, 2000)
  • new Voronoi class to compute Voronoi regions in graphs
  • new graph generators, e.g., randomSeriesParallelDAG()
  • new simple graph algorithms, e.g., makeBimodal() or makeMinimumSpanningTree()
  • and as usual: bugfixes, (a little) code cleanup, and improvement of code quality (like const-correctness)