Graph visualization tool made with godot
using KaHIP for internal graph structure
running on linux
Website: https://pille.iwr.uni-heidelberg.de/~graphvis01/
You can download the project by using:
git clone https://github.com/dnsch/Godot-GraphVis
In order to run the main program, unzip the linux.zip and compile the modified KaHIP following the instructions in INSTALL and README. You might need to make the compile_withcmake.sh executable via:
chmod +x compile_withcmake.sh
assuming you are in the modified KaHIP directory.
After the compilation you can find a libkahip.so file in the newly created directory kahip_source_modified/deploy/.
Copy this file into the unzipped directory linux/lib.
After that, you can run the Godot-GraphVis program in the linux directory via:
./launcher_script
In order to read in a graph, you have to create a folder with at least an adjacacency list file in the Metis file format (see p.8 in kahip.pdf) with the extension .graph and a coordinate file with the extension .coord, storing the x and y coordinates for each node in the associated row. You can calculate the coordinates using a force-directed graph drawing algorithm, such as Fruchterman–Reingold. For the examples found in linux.zip, KaDraw was used to calculate the coordinates in the respective .coord files.
Additionally, if the graph contains a partition file with the extension .prt, which stores the partition id for each node in the associated row, or if it contains a vertex set file with the extension .vs, which stores a 1 if the node is in the vertex set and 0 if it is not for each node in the associated row, these can also be displayed. For the examples found in linux.zip, KaHIP and VieClus were used to calculate graph partition and graph clustering files (which both can be read in as a .prt file) and a random walk algorithm to solve the Minimum Feedback Vertex Set problem to an optimum was used to calculate the vertex set files.
The graph folder then can be read in using the File button on the top left of the gui.
After that, the graph is displayed in white and an additional Options button appears next to the File button. Using the Options button, the appearance of the graph can be changed. This includes coloring edges and nodes by partitions, clusters, and vertex sets, changing the size of edges and nodes, changing the size of nodes based on their degree and coloring the edges based on their length.
Clicking the Camera icon next to the Options button creates a .png that is stored in the directory of the executable.
If you want to use the GraphVis module in Godot, place the GraphVis module in the modules directory which is located in the root of the Godot source and follow the instructions for compiling (Compiling for X11, Custom modules in C++). In addition, you have to copy the libkahip.a from kahip_source_modified/deploy/ (after compiling it) to usr/lib via:
cp libkahip.a usr/lib
assuming you are in the kahip_source_modified/deploy directory.
This allows you then to call the modified kahip source and additional functions found in graphis.cpp as a class within the Godot editor by using the constructor GRAPHVIS.new(). For further information on using custom modules have a look at Custom modules in C++. You need to edit the SCsub if you want to include more functions from modified KaHIP.
Alternatively, you can edit the existing Godot source code for this project.