C graph library for complex networks research. Graphs are stored in-memory in the most compact form possible, hoping to be the most efficient and parallel implementation available.

The following example is available in example/printing.c, and showcases how to create a Barabási-Albert network and plot it as an SVG.

#include <stdbool.h>
#include "graph.h"
#include "graph_model.h"
#include "graph_layout.h"

int main(){
  int n = 100, k = 4;
  // Creates a graph using the Barabási-Albert model
  graph_t *g = new_barabasi_albert(n, k);
  // Position of each node
  coord_t *points = malloc(n * sizeof(*points));

  // Styles to be printed
  int radius = 5, width = 1;
  circle_style_t circle = {
    .width = width, .radius = radius, 
    .fill = {255, 0, 0, 255}, // color is a 4-tuple
    .stroke = {0, 0, 0, 255}
  };
  path_style_t edge = {
    .type = GRAPH_STRAIGHT,
    .width = width,
    .color = {0, 0, 0, 255}
  };

  // Places vertices in shells, where higher degree vertices are 
  //internal and lower degree are external.
  int size = (int)
    graph_layout_degree_shell(g, radius, true, points);
  
  // Plots graph at file barabasi.svg
  graph_print_svg_one_style(
    "barabasi.svg", size+1, size+1, g, points, circle, edge);
  
  delete_graph(g);
  free(points);
  return 0;
}

The current version is 0.1 - id est, alpha stage, when API (most surely) will break.

Code is factored in modules as follow. Each module has corresponding archives

• src/module.c
• include/module.h
• test/test_module.c
• doc/module.tex

Functions related to searching and sorting.

Statistical-related stuff. Needs a major refactoring.

Array list implementation for ints, with support for sorted operations that allow O(log n) searching and insertion.

Hash set implementation for (positive) ints, with O(1) insertion, and O(n) deletion.

Basic operations for creating and populating graphs with a fixed number of vertices. Edge deletion is not supported, and needs some refactorings to support directed graphs properly.

Several interesting graph metrics, such as degree correlation, centrality and geodesic distance.

Layouting and printing graphs into SVG files.

Common models in Complex Network research: clique, random, scale-free and small-world included.

Information propagation models in networks: SI, SIS, SIR, SEIR and Daley-Kendall.

Game theory module, with Iterated Prisioner Dillema implementation.

For more information, please check the documentation, available in doc/main.pdf.